I am quite a newbie on openMV and doing a project using a thermal camera with YOLO. To create the desired thermal camera, I am using the OpenMV Cam H7 R2 + FLIR® Lepton® Adapter Module + FLIR Lepton 3.5. In the OpenMV ide, I was getting the desired thermal image. To work with YOLO, I needed the camera in UVC mode. So, I flashed it with uvc.bin, following this git repository (target_name= OPENMV4). But by default, the camera is outputting a gray image instead of a thermal image.
I also checked with Guvcview with different camera outputs. But I didn’t get the thermal image. So then I found this video of kwagyeman where he got the thermal output using RGBP-16-bit RGB 5-6-5. But I am getting the following camera output.
- Could it be a firmware-related issue? If so, how can I solve this issue?
- How can I get the thermal image by default instead of the gray image?
I don’t believe UVC supports RGB565, so you need to it the way bytes are generated in that firmware to map RGB565 to RGB888. Please look at the main.c file for what you need to do.
Thank you for your quick response.
I found one of your youtube videos where you used RGBP-16-bit RGB 5-6-5 as camera output to show the thermal image.
Did you map RGB565 to RGB888 in main.c to get the above thermal output? If so, then only updating main.c file be enough to get the desired thermal picture?
Besides, how can I change the default camera output gray image to a thermal image?
Dear friend the output of this sensor is agreyscale image. then you are free to convert it with image class like rainbow, ironbow… Please check the examples on ide and you will understand…
Thank you for your suggestion.
I am talking about the camera output in UVC mode. In the UVC mode you can’t use openMV ide.
Hi, yeah, that might have worked with UVC a while back. We don’t test it constantly. But, if you download our build environment it shouldn’t be hard to fix the code for that.
What did you mean by build environment? Can you please provide me the repository link?