My new Hikmicro Pocket 2

[kultakala]

Hi,

I have a 6 year old Flir E4 which i succesfully hacked to the E8 specs and features and its a quite good device especially for the price.
Usually i use it to look for isolation of walls and windows...  heating pipes or electrical stuff.
And of course just for fun

But there are some things i dont like, for example the size which does not fit in a pocket, lack of video recording and the limit to display at only 9 Hz.

So i was looking for an alternative and tested a HTI HT-301 and had in mind the Flir C5.
The HT-301 delivers some good images but i didnt like the handling of the camera/smartphone combo and the manual focus.
Also the software was okay but had it quirks.
The C5 otherwise is pocketable but also lacks video recording, has a low resolution and also 9 Hz limitation.

Then i found out about the Hikmicro Pocket 2.
Its also pocket size (14 x 8.5 x 2 cm), has a 256x192 sensor with NETD < 40 mK, wide field of view (50° x 37°), has a 3.5" touchscreen, 16GB internal storage and has builtin WiFi (2.4 and 5 GHz) and Bluetooth (Audio only).
It can record video with 25 Hz and audio.

I received it last week and all i can say is that it is exactly what i was looking for.
The startup time is about 10 seconds, the display has good colors and brightness.
The menus are intuitive and there are a lot of settings. It has an optical camera with fusion display and configurable distance between 30cm and 3m and above. There is also a macro mode but i couldnt see any benefits using it.
The software for Android (Hikmicro Viewer) is also very good and stable.
Its possible to remote control the camera via WiFi, transfer the images and videos and its possible to edit the images afterwards and change everything from color palette to emissivity and set new measurement points, etc.

Overall i am very happy with the camera 

I was not able to put the images between the text, is that possible ?
The images show the camera itself, electric kettle, mug before and after editing, the pulldown menu and display with all overlayed info enabled and disabled.

[jeremy]

Looks great! What did you pay for it? And where?

[kultakala]

It was about 650 Euros at a german dealership.

[Fraser]

That looks to be a neat little camera. Thank you for sharing your experiences with it 

Fraser

[svgurus]

can you post some pictures comparing e4( and this thingy? their fov is almost identical

[Someone]

Lets keep all this pocket camera goodness together. I bought one as had a pressing need for some higher resolution imaging, and have been impressed so far. There was no need to perpetuate the bulky format that thermal cameras traditionally had and this dense slab is really an improvement in usability.

The not so good:
While it has a shutter for online calibration, the residual fixed pattern error is pretty distracting in smooth areas.
There is a macro lens option but its not appearing in the retail channels.
Some features are blocked when plugged into a USB host.
Latency to the screen is high enough to be difficult for handheld use (video frame rate is fine).
The shutter button is way too easy to accidentally press.
It has a soft-touch rubberised finish (have had similar surfaces go bad too fast for a tool).
Captive battery.

The good:
its cheap! (found a bargain for $400 USD).
USB C charge and photo offload.

So without being able to buy an OEM macro lens, I had to DIY something. Pictured below is a small folded metal can (like an EMI can but with 1mm thick sheet stock) that locates and sits on the casing protrusion around the lens(s), it has one of the cheaply available ZnSe ⌀=20mm 𝑓=50mm plano-convex lenses sitting flat which produces a good working distance and field of view. Lacking a USAF resolution test chart to hand, you'll have to put up with a cheap PCB ruler instead. It picks out the 0.5mm pitch pads pretty clearly, and easily shows 0402 components, field of view is 30x40mm with the macro lens in place.

Accuracy? who knows, but it works and will be a companion on the bench from now on.

[markb1980]

I have created this closeup-lens holder: https://www.thingiverse.com/thing:5397821
If you prefer a ready made holder - I guess Hikmicro got inspired by us: https://www.hikmicrotech.com/en/accessory-detail/23

Results are as expected very good. I love the capability of the Pocket 2 to take videos... The slim design and low weight allow me to mount the camera on a flexible arm so that I have the hands free when I am working. That make it my favourite IR camera for electronics testing.

[markb1980]

Brewing coffee and printing the closeup-lens holder as thermal video:
(just a pitty the video is not radiometric but to expect that would be a bit to much at this price-point)

https://youtu.be/HBQAI6GyIa4
https://youtu.be/sI09LponG7c

If you need radiometric data you can use the timelaps feature and take 1 image / sec.

[Someone]

Not dissimilar to the above by markb1980, but accidentally finding a dodgy (very cheap moulded strain relief banana) cable joint. Those connections are carrying a symmetric current!

[c00kemon]

@kultakala
Thanks for the review.
Can you take several thermograms on pocket 2 and E4 under the same conditions for comparison? It would be great if you take the facades of buildings as examples.
And what is the accuracy of Pocket 2 at ambient temperatures closer to 0 C and below?

[Pfriemler]

Got my Pocket2 a few weeks ago as a replacement for my not repairable FLIR Gen 2 Android mobile plug-on. I am very happy with it. Thermal picture qualitiy is way better than the poor FLIR - as expected.

But there are were things bothering me:
a1) the "fusion" is not that qualitiy known from the little FLIR. The slider in the app was easy to use to align optical and thermal cam (fusion contures). On the Pocket you have to use a scroll list on the touchscreen.
a2) my Pocket has not only a horizontal offset varying by distance to the object, but also vertical. I could live with that, if I can correct this in any way but
b) there seems to be NO WAY to realign optic & thermal pics afterwards. The HIKMICRO Analyzer PC software has nothing like that.
c) with the FLIR PC software I could blend both pics in adjustable transparence, very useful to show heat pipes in floors etc. Again: NO way here. For that few times I need that I have to export both pics and blend them externally with GIMP or something like that 

The good thing: The thermal resolution (pixels and delta) is really impressive so I do not really have the urgent need to blend both pics. 

No experiences yet if used at low temperatures, sorry.

edit: upgraded firmware and pc software solved b) and c), txh to hints here...

[markb1980]

a1 + a2) can be done in the PC software.
b) update to the newest version (if you select fusion in the dropdown in the analyze view there is a point "adujst fusion alignment" in the dropdown)
c) same for blending - use HMA 1.2.0.3!

Also update the Pocket2 firmware for increased sharpness and better thermal contrast as well as fixing some bugs with the date and time.

[Pfriemler]

well, that was too easy ... 

honestly ... I got the cam just a month ago, checked available firmware updates (there were no) and missed the blending feature in live view at the cam (a thing the FLIR does not at all - at least until my device died) as proved in the gun style HIKMicro and many other - and, tataa - there it is! with the firmware 5.5.29, just released a week ago. And a portrait mode, too. Damned banana ware!  . Even fusion seems to be better than before ...?

And with the PC software: I got a German (!) 1.1.somewhat just a few days before the release of the 1.2.0.xxxx (which turns out as a 1.2.0.3 if installed). And - yeaah - there are blending and image position correction just where I expected them before. Works fine.

Meanwhile, I found out that the vertical alignment issue is more on the left, looks like one of the sensors is tilted (I think the thermal is).

And: I was not able to connect my Pocket 2 to my MS Surface Pro to get the firmware file into device's root (I had to consult youtube that only the digicap.dav is to be copied - found no written hint neither at the website nor the manuals    ) - the cam draws too much current from the USB and was rejected to connect. I charged it and it took 1.8A (in words: one point eight ampere) from the wall plug. Nice to know that the cam can charge in that fast way, but what if I have to connect to get the pics from it? 

Nevertheless: with the fresh upgrades really more better.
I miss a (live) viewer for the PC - the Android HIKViewer works fine.

[markb1980]

Even fusion seems to be better than before ...?

Yes that's true and thermal only also improved a lot in situations with a narrow span!

The PC anlysis software got also new tricks like the focus mode palettes and color alarms... I suggested them to add the same pallete as I have used in my profile picture and I guess the focus mode palletes are the way how the developers did understand my suggestion - at least you can decide so if you want to gray out above, below or ouside a certain temperature / temperature-range.

not the same as a palette but basically you can controll it this way even better.

And with the PC software: I got a German (!) 1.1.somewhat just a few days before the release of the 1.2.0.xxxx (which turns out as a 1.2.0.3 if installed). And - yeaah - there are blending and image position correction just where I expected them before. Works fine.

I am an author of a beginners guide to thermal imaging and I have suggested that to hikmicro a while ago. Great they are listening to clients and people as me. I have also a thermal imaging group in facebook and I am in contact with many thermographers around the globe to discuss new cameras and new functions...

Meanwhile, I found out that the vertical alignment issue is more on the left, looks like one of the sensors is tilted (I think the thermal is).

I would talk about that with my reseller.

I was not able to connect my Pocket 2 to my MS Surface Pro to get the firmware file into device's root (I had to consult youtube that only the digicap.dav is to be copied - found no written hint neither at the website nor the manuals

Before there was a manual in the ZIP-file - maybe they forget to include it for this update.

The old firmware was also not adjusting the time when switched off. So I had to set date and time each time I power the camera on. (Very annoying) But that was fixed in the latest update as well.

Nice to know that the cam can charge in that fast way, but what if I have to connect to get the pics from it?

My office PC has no issues with the USB connection but you could try to use bluetooth or WLAN. If not write them a email - they usually respond to issues quite fast. I recently asked them to include one more color palette.

Nevertheless: with the fresh upgrades really more better.

[Pfriemler]

Nice to read that they are interested in suggestions. That's what we thougt here about MicSig scopes ... turns out that it is only to very special persons.

... looks like one of the sensors is tilted (I think the thermal is).

I would talk about that with my reseller.

The question is: is it that much to return it? We are talking about 4-8 cm in a 5m distance. I try to make a demo pic...

Before there was a [upgrade] manual in the ZIP-file - maybe they forget to include it for this update.

They did. That was the place where I expected it. I have a device which expects a zip file instead of the pure binary, that's why I looked for instructions.

The old firmware was also not adjusting the time when switched off.

My Pocket came with the .25 and all files were dated correctly.

My office PC has no issues with the USB connection but you could try to use bluetooth or WLAN.

My home PC do not care about USB currents, too. But I blowed a fuse on the mainboard by an accidental short circuit, leading to a dead USB 3 port.
Both things should never happen. Many devices cut their current to 500 mA when there are data on the lines.
I had the hope that there's a FTP or something like that, but there are only ports 80, 443, one in the 5xxx (streaming) and 8000. I love this on my MicSig Scope, so I use my phone to transfer the screenshots to my home server via android ftp client without the need to turn on the PC at this time. That would be nice here, too. That's worth an email, I think...

[markb1980]

My home PC do not care about USB currents, too. But I blowed a fuse on the mainboard by an accidental short circuit, leading to a dead USB 3 port.
Both things should never happen. Many devices cut their current to 500 mA when there are data on the lines.

... my Pocket2 take 0.46A when hooked up to the PC and 0.82A when hooked to a 1A charger.

So maybe your model has some issue. I will ask another Pocket2 owner if he can confirm that maybe...

[Pfriemler]

... my Pocket2 take 0.46A when hooked up to the PC and 0.82A when hooked to a 1A charger.

As it should be.

So maybe your model has some issue. I will ask another Pocket2 owner if he can confirm that maybe...

Would be fine. But in my experience, the weather outside is more predictable than how much mA a (charging) device on USB will take. With my mobile, that varies from 0,8 to 1,8A - cable, usb output voltage and behavior - all at 5V, no PD, no QC. The mentioned wall plug is a good one and is stable on 5.2V @ 3A. I have a bunch of USB voltage/current meters and will check that next time the battery of the Pocket is at least 30% down. I do not think this is an issue, but we will see.

[c00kemon]

Can you post some sample pictures after the update?  And how the camera behaves in terms of measurement accuracy at temperatures closer to 0 or below.

[markb1980]

Hi. This are images from an old and a new report with the newer processing. I will try to get a sample of a narrower span image but I am not sure if I have one stored in the office.

[markb1980]

That's a 2°C temp. span and it's rendered great. The old version was much more grainy and much lower in contrast even with a bigger span (see 2nd image).

[c00kemon]

Thanks for the image examples.  Amazing image enhancement, more detail in pictures.  It looks like a super resolution algorithm.  And at what price did you buy your camera?

[markb1980]

To be hornest - I got from Hikmicro a M10, B20 and Pocket2 to test for the 2 books I have written about thermography. When I decided to keep the Pocket2 I got a surprisingly great deal. I guess they sell me one for aprox. the production cost so I will not disclose that.

I think the 550 EUR + Tax they charge for the camera are very fair. Compared to a C5 (I own also a Cat S62 which has the same sensor) the Pocket2 is really great in terms of price and performance. But I had also criticized them for the softer image rendering compared to GuideIR or InfiRay multiple times in my FB group and some blogs and for the grainy thermal images with a very narrow span.

I was in contact with then for a while and I shared samples from the other cameras I own (GuideIR PC210, InfiRay C210 Pro and P2 Pro) and I have tryed also various image editing tools and techniques to "improve" images and I have shared that results too. So I am also very pleased to see improvements and new functions in the software because I was talking with them about these things many months ago. I was also in contact with a few Level 2 and Level 3 ITC certified thermographers for the book and we where talking about some of the critique points I had in some of the cameras to get a better idea what can be improved and what is a limitation of the technology or the price range in which the cameras are placed.

When looking back a few years where cheap thermal imaging started with a VT2 (aprox. 32x32 pixels) and where we are now just a little over 10 years later I am pretty impressed.

[c00kemon]

Wow. I also have many thermal cameras such as Infiray c200, P2, Guide Pc210 and the reference for me is a hacked Flir e4 wifi.  I, too, all the time something does not suit me in the software of Chinese thermal imagers and I try to help them, especially Guide IR, I wrote bug reports for them, but so far without much success.  I'm just as surprised how thermal imagers became available to a not so rich person.  Give me a link to your community, it will be interesting for me to read new information.

[markb1980]

https://www.facebook.com/groups/1816479168568552

Please also share your bug reports and informations with the group. I have also a few people from Flir, Guide, Infiray and Hikmicro in the group...

For now Hikmicro is the best company to deal with in therms of suggestions but Guide was also open to it. Hikmicro really listen to their clients and their distributors.

[c00kemon]

Ok, thanks for the reply.  I finally found the right person with similar interests)

[katzenhai2]

To be hornest - I got from Hikmicro a M10, B20 and Pocket2 to test for the 2 books I have written about thermography. When I decided to keep the Pocket2 I got a surprisingly great deal.

Is there any difference between the B20 and Pocket 2 beside the better screen of the Pocket 2? I think Hikmicro will soon have to release new handhelds (B-series) that also have this 640x480 touchscreen.

I can only hope that the competition will get up and push new developments so that we experience the same development as in the outdoor area: There are already VOx sensors with 640x480 and <20mk on the market. I would also like to see that at reasonable prices for the measuring range/handheld devices. Sensor resolutions with 384x288 in the B-series and 640x480 in the M-series with lower NetD at the same current prices are overdue.

Or am I asking too much? 

[markb1980]

Is there any difference between the B20 and Pocket 2 beside the better screen of the Pocket 2? I think Hikmicro will soon have to release new handhelds (B-series) that also have this 640x480 touchscreen.

No basically it's the same unit and the same temp. range and the same resolution. The Pocket2 is smaller and thats why I use that one. I can have a Pocket2 always with me in my laptop bag...

Or am I asking too much? 

Would be great. But when you look back a few years you will see that affordable thermal imaging has come quite a long way from Flukes VT02 to a B20 or Pocket2! But I hope also it will progress and a Pocket2 / B20 with 384x288 would be really great!

[Jeremy Morgan]

I was thinking of getting one of these. I couldn't see any reference in the thread as to how well this imager works with PCBs (checking for thermal issues of course). Does it focus close enough?

I tried a Bosch one a few years ago but it was clearly really designed for use at the room/building scale rather than close up.

Thanks!

[markb1980]

Not without a closeup-lens (50.8mm ZnSe lens for a lasercutter). But with such an inexpensive lens and a 3D-printed addon (https://www.thingiverse.com/thing:5397821) you get nice results...

[markb1980]

Without you can get an Idea but pinpoint a issue may be difficult. That a image from a B20 without the adapter.

[katzenhai2]

Hikmicro also have a macro lens to offer:
https://www.hikmicrotech.com/en/industrial-products/pocket-series-macro-lens.html

[katzenhai2]

What I have to complain: It would be better if the images were in PNG format. Then there would be no compression loss since PNG is a lossless format.

[markb1980]

The JPEGs are radiometric - so they contain all measured values and the software can alter and analyse and newly render them als needed. So consider a radiometric JPEG as something like a RAW image.
The images are also upscaled to 640x480 so even PNG would not loop perfectly fine.

As far as I know it would be also more challenging to add the radiometric data into the PNG header but I have not looked so detailed into PNG and into JPEG.

[katzenhai2]

The JPEGs are radiometric - so they contain all measured values and the software can alter and analyse and newly render them als needed. So consider a radiometric JPEG as something like a RAW image.
The images are also upscaled to 640x480 so even PNG would not loop perfectly fine.

Thats true. But then you always have the need to forcibly load the photos into the software. The photos as they come out of the camera are often sufficient enough without tweaking afterwards. If you then have compression artifacts (which you would also have after saving them with the software) it doesn't look so nice - e.g. for reports.

[ExaLab]

I received this interesting "toy" only a few days ago and the device, overall, seems to work very well.

On my unit, I noticed the following firmware anomaly: regardless of the humidity value set on the thermographic settings, the humidity value obtained by recalling the info field of the taken photos is always equal to 60%. All other thermographic parameters are reported correctly (i.e as setted).

Has anyone noticed this bug?

[markb1980]

What firmware-version do you have?

[ExaLab]

What firmware-version do you have?

V 5.5.29 build 220909

[markb1980]

I have the same issue with v5.5.29 build 221029!
Please report that bug to the support...

[ExaLab]

Many thanks!

In the meantime, the doubt I have is: which humidity value is used by the device to calculate the temperatures, the one in the "settings" menu or the one shown in the "info" field of the images?
May be that the answer to this question could come from the Analyzer software. Anyway, the sensivity to this parameter should be very low...

[markb1980]

I can't tell you that but I have used Flir Thermal Tools where I am pretty confident all calclulations are correctly.
I set the distance to 30m because the humidity is used to compensate for water in the air blocking IR radiation. The difference between 10% and 90% rel. humidity is the difference between 13.5 and 13.6°C!
With a distance of 300m the difference between 10% and 90% rel. humidity is the difference between 12.9 and 13.5°C...

So there is a difference but it's very minor and not important in normal applications. This factor become more important when the distance to the subject get bigger. It also "just" effect the temp. calculation but not the visual representation. I hope that put at least the importance of that factor into perspective.

[Bill W]

I set the distance to 30m because the humidity is used to compensate for water in the air blocking IR radiation. The difference between 10% and 90% rel. humidity is the difference between 13.5 and 13.6°C!

For those temperatures that may be the case.  However when running around 25°C and 30 - 80% RH there is a bigger difference.  Even setting up cameras at 4m (BST cores, for a set output signal to represent a 5°C difference) there was a significant difference 4pm to 8am, 10% of signal.
It went away close up, so was air absorption, not simply the chart 20-25 becoming 25-30 °C as the day warmed up.

[markb1980]

I tryed it in Thermal Studio with an object which is ca. 28°C warm.
I set 30m distance and 10%RH vs. 90%RH was resulting in 28.4°C vs. 28.7°C. The same with 300m distance resulted in 28.8°C vs. 29.9°C.
To reach the +/- 5°C between 10% and 90%RH I have to set a distrance of 1400m!
 

[ExaLab]

In these days I have had the opportunity to test the Hikmicro Analyzer software.
I find that it has a good structure but at the same time it lacks of many basic functions some of which I have already had the opportunity to report to the company support (without receiving any feedback about it...).

Below, what I observed regarding this SW:

1) In "Image Analysis" the "PIP" Display Mode lacks the possibility to resize te PIP frame

2) More, the same "PIP" Display Mode lacks the possibility to remove the "Fusion" effect from the thermal image frame (in order to have a pure thermal image on it).
This aspect also concerns the firmware of the thermal imaging camera (Pocket 2). Even in this enviroment, it would be nice to have the possibility to remove the fusion effect

3) A big problem also affects the three panes on the right side of the Image Analysis window (named "Image Information, Measurements and Parameters respectively). If you try to scroll with the mouse wheel on these windows, if the pointer is above one of the many parameters present, the relative value will inevitably be altered!!!.
You can do enormous damage, even without realizing it, simply by rolling the mouse wheel!
Incredible the fact that none of the development team has noticed this obvious criticality!!!
More, once this damage is done, there's no going back. You have to start from the beginning, reloading the image.

4) Another incredible lack affects the "Report Editing Page". If you try to resize any image, you will lose its aspect ratio!!
There is no way to preserve the image ratio nor to recover it!

It is my opinion that nobody from the development team has ever deeply tested this software...
A software full of gaps and which contrasts with the excellent firmware of the Pocket 2 camera.

Is anyone able to tell me some valid alternative SW dedicated to the analysis of thermal images?

[markb1980]

There is none. If you test the software of Guide or even Infiray they are much worse. Infiray alter for example dates and other Information on the image and Guides Software don't even factor in changes in RAT or emissivity afterwards, etc.

There is nothing better that's why I tell always there is no other cineese brand you can use for prof. work except Hikmicro. The Software works at least and you can get alon these issues. At least you get correct results out of the software which (as sad as it sounds) is the best you can expect if you don't go with Flir, Fluke or Testo...

If you want better Software you have to go with an Flir E6 and pay 4x more or you go with a Testo 868s and hope the Super-Resolution will give you the same image quality as native 256x192 or you go with a Testo 871s and pay 4x more and not just 2x more.

[ExaLab]

Thanks a lot!

I'll try again to report to Hikmicro these lacks / criticalities, hopefully they will be fixed in an upcoming software release...

Question: can the images produced by the Hikmicro Pocket 2 be loaded and processed by the starter version of Flir Thermal Studio or are they incompatible with it?

[markb1980]

I have 4 different cameras and there I have to use 4 different programs. No file-format of any camera can be edited in another software theeen the manufactorers one. Some manufactorers like Uni-T and Infiray have even different software for different cameras!

So the answer is no. All manufactorers use prop. formats.

There are some advanced reporting tools but as SaaS offering and you have to pay some money per month: https://ti-reporter.com/

[ExaLab]

Thanks for your support Mark.

I'll try to make the best use of what I have!

[markb1980]

I can even help you more - if you want I can send you a few sample-images grom Guide and Infiray. After trying there software you will be really happy with what you have 

If you want to try other chineese reporting tools send me your email address as PM.

[rfspezi]

@markb1980

How did you add the Pocket 2 camera to HIKMICRO Studio?
When i add a device, the mask attached is opened - however, what port number, user-name and password do i have to enter?

[markb1980]

I am not sure you can do that. Studio is as far as I know for the network / fix installed cameras

[rfspezi]

Anybody found a suitable screen protection foil for the HIKMICRO Pocket 2?

By the way... the swipe down gesture only works reliably when you do it exactly in the middle of the screen.

[markb1980]

There is non to my knowledge. But there are a few companies in europe which cut screen protectors in each size you want...
All you need is width, height and corner-radius. I have checked one in Germay they would charge aprox. 9 EUR for a fitting screen protector.

[rfspezi]

Ok, so after using my Pocket 2 for a view days, it randomly crashes/freezes! 
The problem always starts with freezing high/low-temperature marker positions and when entering a submenue, the camera stops reacting to touchscreen inputs at all.
I can still hear it making the clicking noise when recalibrating.
To turn it off i have to long-hold the power button.
A really bad experience after a short usage time for such an expensive device.... disappointed about the firmware quality.

[markb1980]

You have probalby buyed the camera at a local retailer. So tell hin that problem. I have heard in my group that someone had the same issue - the retailer replaced the device or did a firmware-update... If not try to update the firmware even if it's the latest version.

I guess that's rather a QC issue then a firmware-issue...

[rfspezi]

I can almost always replicate the freezing when i do the following:

1.) Start the camera
2.) Wait for the first "calibration click" and immediately take a picture
3.) Wait for e view seconds and the min/max temperature markers stop moving (or worse - crash)

I am on the latest firmware version V5.5.29_221029.
Does anyone still have the previous firmware version available or reproduce the issue?

EDIT:
As expected, updating the to the same firmware version again did not change anything.

[markb1980]

Ask the Support or retailer...
It is a know problem and as far as I recall it ist because of a bad installed firmware...

[rfspezi]

Ok... i contacted MIKMICRO support directly but thinking about returning the camera completely.

By the way...
Another strange thing once happened on startup.
The screen was black showing the text "NO VIDEO" for approx. 10 seconds. 

[markb1980]

Last idea would be a factory reset and format of the memory...
I also asked in my group for you - maybe someone have an older FW for the Pocket2.

[rfspezi]

Thank you.

That was exactly what i did some minutes agoe.... no change

[carlob]

Ok, so after using my Pocket 2 for a view days, it randomly crashes/freezes! 
The problem always starts with freezing high/low-temperature marker positions and when entering a submenue, the camera stops reacting to touchscreen inputs at all.
I can still hear it making the clicking noise when recalibrating.
To turn it off i have to long-hold the power button.
A really bad experience after a short usage time for such an expensive device.... disappointed about the firmware quality.

Hello.
I also must confirm this behaviour. Also freezes, the camera works, but you can't do anything with it.  It freezes high/low-temperature marker positions and can't go into any menu.

I searched for this issue and found this forum. And it looks like I registered here a long time ago.

I also notice something. When the camera is frozen, it is also impossible to charge the battery! You can plug in a USB-C plug, but nothing happens. LE Diode is off. Nor red, nor green.
Firmware is upgraded, to the latest version, 5.5.29.

So, I have tried to reset the camera. The Gear wheel icon on the right part of the screen frame, bellow photo album, then Device Settings, then Device Initialization, then Restore, pop up menu asks Initialize, touch OK.

After initialization, I played for a half hour and never was able to reproduce that freezing any more & I took more than 100 photos.
Just once, when I press the shutter button during the calibration process, the camera's High/Low marker freezes for ten seconds. After that, the popup menu went, "capture unsuccessful" and the camera starts to work again normally.

I tried a few times more and I can easily reproduce that 10-second freeze, if I, immediately after booting of the camera start to press the shutter like crazy. It will block capturing for 5-6 seconds and after the "Capture unsuccessful" sentence, it continues to work normally. Also, the camera gave you a pop-up with a "please wait" sentence, so you should wait for initialisation and calibration to finish before you start to capture.

Please try the "Initialize" procedure. It will not remove or erase any photo! I have tried.

Thanks for your posts and experience.
I think this small camera is of incredible value, especially its resolution and 25FPS! For this money, Flir has shamefully 7-9 FPS and low resolution.

[markb1980]

I think this small camera is of incredible value, especially its resolution and 25FPS! For this money, Flir has shamefully 7-9 FPS and low resolution.

Yes but there are other brands like GuideIR with the PC210 or InfiRay with the C200 which offers the same specs for aprox. 400 EUR...
So compared to other cineese brands a Pocket2 or B20 is quite expensive but Hikmicro offers other things like a calibration certificate and by far the best software among chineese brands and a support and local dealer network like flir. Hicmicro is also the only brand which allow me to compensate for a reflection of the clear sky in this price range. etc.

So I am still sure it's the best option among the chineese brands but not everyone would need that. If you don't need to write repots or you don't need to have best possible accuracy for readings you may get happy with other offers. I know a lot of electronic repair guys which are happy with a Uni-T UTi260b or a C200...

Flir is not cheap and will be never cheap. But flir has the 9Hz refresh-rate because of export limitations for example. Flir is also known to have the best software and noone would question a Flir camera in court. So most thermography experts still use flir.

I understand your point but you are not fair here and you overlook also things like R&D costs, manufactoring costs, etc. in different countries.

[carlob]

Yes, but Hikmicro also has an M30 with even better resolution and a fabulous G60 with VGA resolution & 50 FPS!!

It will eat Flir for breakfast, just like DJI destroy all other drone producers. I know that they limit FPS to 7-9 FPS for civil use, but that is their handicap! They (the USA) are killing kids around the world with their bombs, so I'm not sure why limiting something like that. Who would use that feature against the USA, when they use smart bombs to kill children from 10 miles above, like cowards!

But as I said, they can limit, I won't buy! Their loss, not mine!

Yes, Flir has military contracts etc, but in a few years, Hikmicro and Chinese VOx chips will be so great and so sensitive, that they will simply wipe the floor with both Flir & Fluke.

My Pocket 2 easily sees cat paws' thermal footprint in great precision, contrast and clarity. Flir C5 can't see that quality and resolution.

Those brands like PC210 and InfiRay, I can't be able to buy them in my market, so... I have not tried those, but I'm not sure if thermal characteristics and thermal resolution are in reality same as Hikmicro.

I use to watch some videos of Uni-T UTi384H because I would like to have 110k of IR pixels (384 x 288 res) for reasonable money, I watched every video of Uni-T, but their IR meter with that resolution simply can't compare with HikMicro one.

And In G60, the VGA 640x512 chip is incredible with a ridiculous 50FPS!!!

Flir would charge their VGA instrument in 5 digits range, and the first digit would not be "1"

Thanks for your comment.

[markb1980]

Those brands like PC210 and InfiRay, I can't be able to buy them in my market, so... I have not tried those, but I'm not sure if thermal characteristics and thermal resolution are in reality same as Hikmicro.

I would say the C200 and PC210 deliver a bit sharper images. See: https://www.eevblog.com/forum/thermal-imaging/guide-pc210-review/
Or the attached images of the C200 Pro...

At least the last update made Hikmicros images much better. See comparisons...

I use to watch some videos of Uni-T UTi384H because I would like to have 110k of IR pixels (384 x 288 res) for reasonable money, I watched every video of Uni-T, but their IR meter with that resolution simply can't compare with HikMicro one.

I would not see that so - there offering is really great but who need a TIC for almost 1000 EUR without any tool for reporting?!

[markb1980]

That was exactly what i did some minutes agoe.... no change

What's the module-version?

[rfspezi]

Module: V1.0.2.0 build220613
Version: V5.5.29 build 221029

[markb1980]

Someone in my FB group had  older version on his PC. I can email you the file with we-transfer if you want to try a downgrade...
But I was right - it's a known issue with some devices! If you want send me your email a PM.

[carlob]

Mine:

Module V1.0.1.0 Build 220331
Version V5.5.29 Build 221029

[markb1980]

Module V1.0.1.0 Build 220331

... that's the trouble-free version!

I have the same but i ordered one for my gf also and that will probably have the issue. 

[rfspezi]

What is meant with "Module"?
Is that a firmware version or a hardware version?

[markb1980]

What is meant with "Module"?
Is that a firmware version or a hardware version?

Exactly. Version = firmware version and Module = hardware version.
The last firmware works perfectly fine with the older modules but the newer ones have some issues.

[ExaLab]

My unit has the same module of "rfspezi", i.e. V1.0.2.0 build220613.
I hope someone has reported this annoying criticality to the manufacturer in order to have a firmware update soon.

[ExaLab]

Question: does the problem only occur in the instants following the power-on (when the device is probably still initialising...) or can it also occur later, when the camera has been in use for some time?

[markb1980]

I hope someone has reported this annoying criticality to the manufacturer in order to have a firmware update soon.

As far as I know they are working on it but it could nit hurt to report it again and put a bit core preasure on them to spped things up. As more customer complains they have as faster will things go (at least I hope so).

But the same offer applies to you - if you want to downgrade (which sadly also downgrade image quality) send me your email as PM and I send you a older firmware-version...

[rfspezi]

With my device it does not depend on power on time.
It can be reproduced any time.

As markb1980 suggested please also write to HikMicro support complaining about this issue to hopefully get the issue fixed soon.
As for me, i will return the cam and maybe buy it again if they fixed the issue.

[ExaLab]

On my unit V1.0.2.0 build 220613 I'm not able to reproduce the issue.
If I try to take a picture during the calibration cycle, nothing strange happens.
The markers continue to regularly work at the end of calibration time.
The only effect that I observe in these circumstaces is that the camera doesn't save any picture...

Keep in mind that I have never updated the firmware to the last release (my actual version is V5.5.29 build 220909)

[markb1980]

For all experiencing the freezing issue with a Pocket 2 - try that software-version: https://hikmicrotech.app.box.com/s/xtv5qj9qn2izokm237ajo7b89wt6nmrx

It deactivates the rotationg feature but should fix the issue. My contact at Hikmicro told me they work on a new update which will solve that and this is a intermediate solution to get the cameras to a working state again.

[ExaLab]

I usually use this camera with a Windows 10 PC and despite everything works properly, every time I connect the camera to the PC via USB I receive a windows notification from "autoplay" with the following message: "There is a problem with this unit. Scan the drive and fix it".
Of course, the scan of the unit doesn't find any errors.
I also tried reformatting the camera memory but the "problem" persists.

Considering that everything works fine, to avoid this strange message, I simply inhibited the notifications from "Autoplay".

Question: have any of you noticed this behavior?

[ExaLab]

A useful graph (specific to Pocket2) which allows you to evaluate the temperature error associated with the acquisition of hot (or cold) spots of relatively small area compared to their surroundings. The size of these areas (square in the model I hypothesized) is expressed in effective pixels of the sensor.

[Pfriemler]

... windows notification from "autoplay" with the following message: "There is a problem with this unit. Scan the drive and fix it".
Of course, the scan of the unit doesn't find any errors.

These are two different things. "autoplay" rules what to do if a storage device is connected. I always deny any actions, btw.

The "scan and fix it" error I have with any kind of storage devices (e.g. usb sticks) that are not dismounted manually before disconnecting them. The Pocket 2 is mounted like any usb stick (at least in my Win10). Try to "safely remove" the Pocket2 before you unplug it. This should avoid this error messages in the future.

[Pfriemler]

Hikmicro also have a macro lens to offer:
https://www.hikmicrotech.com/en/industrial-products/pocket-series-macro-lens.html

Got mine two weeks ago. A lot of bucks for a few cheap items, but it works fine.

[ExaLab]

I know the role of the Autoplay service. I too usually inhibit its actions when I connect a USB device.
I mentioned it for the simple reason that the notification I receive is managed by this same service, nothing else...

Naturally, in my case, this annoying message is also present with the "Safely Remove" procedure.

What is the FW version of your Pocket2?

[markb1980]

Got mine two weeks ago. A lot of bucks for a few cheap items, but it works fine.

... it's not about the parts - you can easyly make your own. For the original one there is a setting in the firmware (macro mode) which compensates for the lens and allow you correct temp. measurements.

That's why it's basically more expensive. If you don't need that, then you can get along with a DIY one also.

[Pfriemler]

I know the role of the Autoplay service. ...

IMHO it is two things: mount a mass storage and look what to do with it. But this is not important to here.

Naturally, in my case, this annoying message is also present with the "Safely Remove" procedure.

Damned.

What is the FW version of your Pocket2?

V5.520 build 221029

Nice... just to proof my ideas I just plugged the cam to the PC - and got the same error. But... it disappears after a few seconds and all is fine.
Plug out without removing, plug in - error, stays in the info center. Scan and fix - ready in a few seconds.
Remove safely, plug off, plug in again - no errors.

[ExaLab]

In my case the message always disappears after a few seconds and never affects the notification center.
Even after a "scan and fix" + "safely remove" it reappears regularly.

Maybe this behavior is related (I hope...) to the usb driver associated with the FW installed on my Pocket2 (V5.5.29 build 220909)

[santasemilla]

Hi everyone.

I just found this thread about the Pocket 2. I'm also having this issue being on Module: V1.0.1.0 build220331. I can reproduce the freezing bug when the camera goes off after 10min of inactivity, and then turning it back on.

I already dropped a claim to Hikmicro support with no replies up to date. I just checked HM's website and they removed the firmware links (https://www.hikmicrotech.com/en/industrial-products/pocket-thermal-camera/). Also as for today at the Download section (bottom of the page) the User Manual link states the 5.5.45 version. ...maybe we are some days away to a fix.

[ExaLab]

...maybe we are some days away to a fix.

I hope! The firmware has been removed from the site for over a month...

[ExaLab]

Another graph (specific to Pocket2) closely related to the previous one I posted above.



It highlights the temperature error that arises when passing between two large isothermal regions at different temperatures separated by a straight line.
As you can see, for a delta T = 60 degrees, in order to have an error on the acquired temperature less than one degree it is necessary to move away from the boundary between the two areas by about 4 pixels.



In both scenarios (this and that of the previous graph in which hot (or cold) spots were analysed), the temperature errors encountered (which are more or less common to all thermal imaging cameras) are mainly due to the same phenomenon: the static crosstalk between adjacent pixels of the thermal sensor (in part, the phenomenon can also originate from the optical group...)

Obviously, the two scenarios considered are rather "extreme" and normally not found in typical daily acquisitions, characterized by rather soft temperature transitions...
With them I have only tried to highlight a typical error that can occur in the acquisition or analysis phase if one comes across situations similar to those described above.

Be careful! The error I discussed is also present in the case where the two areas have exactly the same temperature but different emissivity!
For example, a stainless steel isothermal plate (emissivity = 0.08 - 0.10) partially covered with insulating tape (e.g. Scotch Super 33+ emissivity = 0.95).
In this case, by setting the emissivity to 0.95, and aiming at the area covered with the tape, we will be able to read the actual temperature of the metal plate.
With the same setting, if we target the uncoated area (i.e. to the stainless steel), we will read a considerably lower temperature.
This delta T is the temperature step seen by the sensor and this is the delta T that must be used to quantify the errors I mentioned above.

[rfspezi]

...maybe we are some days away to a fix.

I hope! The firmware has been removed from the site for over a month...

I am starting to get disappointed about the quality/customer care of Hikmicro.
Fixing such a trivial (yes it is!) issue which is really annoying should not take months!

[santasemilla]

...maybe we are some days away to a fix.

I hope! The firmware has been removed from the site for over a month...

I am starting to get disappointed about the quality/customer care of Hikmicro.
Fixing such a trivial (yes it is!) issue which is really annoying should not take months!

Yup. I'm sorry Hikmicro but I just posted a message in several of your IG posts asking for a solution to firmware problems in the Pocket 2. It's been more than a month that I opened a customer support ticket and no answers whatsoever. I know it is not polite but giving no answers is not polite either.

[ExaLab]

Nunzio vobis gaudium magnum!
Habemus Firmware!

V5.5.45 is finally available on the site

[Pfriemler]

V5.5.45 is finally available on the site

Which? https://www.hikmicrotech.com/en_us/support/download-center/firmware-download/ lists 5.5.29 from Nov'22

[ExaLab]

https://www.hikmicrotech.com/en/industrial-products/pocket-thermal-camera/

Scroll the page or click on Firmware

[ICF2022]

Does anyone know if the new Firmware (V5.5.45) fixed the freezing bug? Or any other noticeable improvements with this version? Thanks.

[rfspezi]

Fixed it as far as I can tell by now.

[ExaLab]

On two different occasions I reported some bugs afflicting the device, not least the failure to record the correct ambient humidity value set.
Well, after months and months of "work" on the firmware these gentlemen (i.e. Hikmicro) weren't even able to solve this trivial problem.

My feeling is that they really don't care about user bug reports and this new firmware (V5.5.45) is proof of that!
All the problems I scrupulously reported and described in detail are still present!!! I really can't believe it!

Total disappointment!

[ExaLab]

Can any of you who have installed version V5.5.45 give me confirmation of the following bug:

  1) In the settings, set a humidity value to any value (except 60%)
  2) Take a shot
  3) Read the humidity value of the taken shot (e.g. clicking info on the taken shot). It will be wrong (you will always read 60%)

[ICF2022]

Can any of you who have installed version V5.5.45 give me confirmation of the following bug:

  1) In the settings, set a humidity value to any value (except 60%)
  2) Take a shot
  3) Read the humidity value of the taken shot (e.g. clicking info on the taken shot). It will be wrong (you will always read 60%)

Bug confirmed in the latest update. I set the humidity value to 30% and the info of the shot taken was 60%. My guess is HIKMicro won't fix this issue but hopefully I'm wrong and they come out with another update to fix it.

[ExaLab]

Bug confirmed in the latest update. I set the humidity value to 30% and the info of the shot taken was 60%. My guess is HIKMicro won't fix this issue but hopefully I'm wrong and they come out with another update to fix it.

Thanks a lot for your kind feedback.
For the third time I informed Hikmicro of this problem and after begging them to pass my report to a product manager, maybe soon we will have a buid that fixes this bug.  In the meantime the last FW has been removed from the site...

Never so difficult to communicate with a company!

[birdy2man]

Hello,

I have now bought a Pocket2 after reading many tetst.
Unfortunately, my camera also still has the V5.5.25 as firmware. On the homepage of Hikmicrotech is currently no firmware available.
Does one of you have the V5.5.45 available and can provide me with it?

Thanks

[RainbowDash]

Firmware 5.5.45 seems to be available here:
https://www.hikmicrotech.com/fr/industrial-products/pocket-thermal-camera/

[pk3]

Hello,

what are differences between Hikmicro B20 and Pocket 2 regarding precision?
I know, this question was asked and answered several months before, but when we look at these 2 cameras, there are some differences - at least when comparing specs:
1. B20 has thermal range from -20°C - +550°C, Pocket 2 only -20°C - +400°C. My local seller explained that Pocket 2 is smaller and cannot be thermally compensated for temperatures above 400°C.
2. B20 after latest firmware supports Super IR resolution 512 x 384 pixels, Pocket 2 does not support Super IR yet

Does it mean that bigger body of B20 allows to do more precise measurements? Does it affect also thermal noise?

Thanks.

[Pfriemler]

sorry, can not contribute to pk3's questions.

Meanwhile, 5.5.49 is available (dated 25-07-2023).
only minor changes: Hotspot offers QR for a quicker connection via WLAN.
Optimizations
Pocket Series handheld thermal cameras have been optimized to improve your experience.
1 Added instant snapshot thumbnail after capturing in live view
2 Improved on-screen keyboard experience.
3 Display order of blending mode adjusted.
4 Save Visual Image: The default visual image resolution is adjusted to 8 MP in this version.
5 Swipe-down menu: Tap and hold Wi-Fi, Hotspot, and Bluetooth icon in swipe-down menu to enter corresponding configuration interface.

to 1: my actual 5.5.29 had it already ...?

I missed to download 5.5.45. I like to see the news/opmimizations in this version. May someone look into the pdf ...?
And ... firmware is declared to fit botch Pocket 1 and Pocket 2. Hmm?

[rfspezi]

Can someone confirm if this is yet another bug of the Pocket 2 or if i get something wrong here:

*) First, I set the temperature range (color bar) manually to e.g. 12,6...31,1°C and lock both limits.
*) Second, i take a picture shown in img1.jpg
*) When i now point the camera to another scene, i would expect the camera to show the same color for the same temperature. However, when i pan the camera slightly to the left side of the first scene so that other temperatures come into view, the colors of the entire image change (compare the corresponding surfaces of the first and second image)

Firmware: V5.5.49 build 230628
Module: V1.0.2.0 build 220613

Edit:

To me this clearly does not work as intended and is absolutely unusable.
I really wonder why nobody else yet has found that issue!

By the way... they even mention in their user manual that the color range should adapt/be limited to the manually set temperature range (page 10).

[electr_peter]

I don't have Pocket 2 to try, but this should not be happening in manual (locked) temperature range mode. Same objects should maintain their color as camera pans around otherwise manual range mode is pointless

Text from manual indicates that temperature range locking should work properly, with objects below/above range mapped to other colors (possible gray colors instead of regular palette). Some IR cameras work this way.

EDIT: looking at the images closely it seems that manual set range [12.6, 31.1] is ignored and scene min/max is taken as actual range ([17.2, 25.2] & [14.4, 26.5]), i.e. it is auto ranging.
Try setting range to [15, 25] and take a photo of something hot (>50C). Maybe manual range will kick in this case.

[rfspezi]

It is definately a bug in the firmware.
Setting the colors to linear mode does not solve the issue.

Sadly i have the same "is it a bug?" discussion going on with Hikmicrotech support as last time when i reported the camera-freezing when taking a picture immediately after startup bug.... Ufffff....

Hope this time they admit the bug sooner and fix it faster than last time (took several moths to get the bugfix!).

Hope dies last.... but it dies...

[ExaLab]

The camera algorithm locks the color palette to the manual selected range of temperatures only if the temperatures range of the image includes the manual selected range (the typical use of this function responds to the need to enhance the contrast of a small part of the image).

Therefore, in your case, to have the same color/temp reference for any of the two images you have posted, the manual range you have to set should be 17.2 - 25.2 degrees centigrade.

If the max of the image is below the max of the setted range and/or the min of the image is above the min of the setted range, the camera algorithm compresses part of the color palette in order to enhance the contrast of the image. This behaviour, on the one hand is positive as it enhances the temperature differences that would otherwise not be visible, on the other hand it is negative as it does not allow comparisons between different scenarios (as you rightly observed...)

To summarize, currently, the locking of the color palette to the manually setted temperature range is only partial.
Personally I believe that this behavior responds to a specific design choice rather than a firmware bug.

For particular needs such as the one you have explained, you could ask the manufacturer to include in the menu the possibility of activating (as an option) the full lock of the color palette.

At the time, I struggled quite a bit to get the bug relating to the set humidity corrected and on several occasions I also asked to include the possibility of removing the "fusion" effect from PIP images, without any success...

[Toumal]

Does anyone have any info about the built-in webserver of this device? I see ports 80 and 443 open (in addition to 554 and 8000) but all I get is a redirect to "/index.asp?randnum=XXXXXX" - which then gives a 404 error... Is this just some weird leftover or is there something that can be done with that?

Oh and... port 554 appears to be RTSP, does anyone know the RTSP path?

EDIT: I did some digging.

The API is at http://<deviceIP>/ISAPI/

You need a session ID which you can get via a request to /ISAPI/Security/sessionLogin/capabilities?username=admin&random=10030310

You can get the RTSP stream at rtsp://<deviceIP>:554/Thermal/Channels/2/ for example, but it's not plain RTSP apparently and won't play in VLC directly - you only get the first frame of the normal camera.

You can set the stream parameters by sending an XML like this:

PUT /ISAPI/Thermal/channels/2/thermometry/basicParam

Code: [Select]

<?xml version="1.0" encoding="UTF-8"?>
<ThermometryBasicParam version="2.0">
<id>2</id>
<enabled>true</enabled>
<streamOverlay>false</streamOverlay>
<pictureOverlay>false</pictureOverlay>
<temperatureRange>-20~150</temperatureRange>
<temperatureUnit>degreeCentigrade</temperatureUnit>
<temperatureCurve>close</temperatureCurve>
<emissivity>0.97</emissivity>
<distanceUnit>centimeter</distanceUnit>
<specialPointThermType>centerPoint</specialPointThermType>
<distance>150</distance>
<reflectiveEnable>true</reflectiveEnable>
<alert>45.0</alert>
<alarm>55.0</alarm>
<showTempStripEnable>false</showTempStripEnable>
<thermalOpticalTransmittance>1.000</thermalOpticalTransmittance>
<externalOpticsWindowCorrection>20.0</externalOpticsWindowCorrection>
<AlertOutputIOPortList></AlertOutputIOPortList>
<AlarmOutputIOPortList></AlarmOutputIOPortList>
<alertFilteringTime>0</alertFilteringTime>
<alarmFilteringTime>0</alarmFilteringTime>
<displayMaxTemperatureEnabled>true</displayMaxTemperatureEnabled>
<displayMinTemperatureEnabled>true</displayMinTemperatureEnabled>
<displayAverageTemperatureEnabled>false</displayAverageTemperatureEnabled>
<thermometryInfoDisplayposition>top_left_of_screen</thermometryInfoDisplayposition>
<emissivityMode>customsettings</emissivityMode>
<displayCentreTemperatureEnabled>true</displayCentreTemperatureEnabled>
</ThermometryBasicParam>

When you first connect to the device in Hikmicro Studio you can set a password. I haven't checked how the flow of that works, but I did notice that vlc asked me for username and password after that when I tried connecting to RTSP. The official documentation on their ISAPI is only available to approved partner companies.

[ElectronMan]

The default username and password is:
admin / HMsc2020

You can enable SSH (don't get too excited, it is a very limited psh shell) using:

Code: [Select]

curl -X PUT -v --digest -u admin:HMsc2020 "http://<IPADDRESS>/ISAPI/System/Network/ssh" -d "<?xml version=\"1.0\" encoding=\"UTF-8\"?><SSH><enabled>true</enabled></SSH>"
There are a lot of options you can play with via the ISAPI interface, and there are a lot of PDF files describing commands online. Streaming the visual camera to VLC via RTSP works if you use rtsp://admin:HMsc2020@<ipaddress>.

I am sure there is a specific URL that will work for the thermal as well, but haven't played with that much yet.

Edit:

For the screen view including thermal:
rtsp://admin:HMsc2020@<ipaddress>/Streaming/Channels/201

[An0v0meo~]

you can try "admin" "abcd1234", hikvision use this user name and password, maybe it can works in hikmicro too