Kristoffer For the colorcamera.preview output there are settings for colororder and interleaved. I would like to take still images using the colorcamera.still and try to time them for each object passing by (maybe 2-5 fps). Are there settings for colororder and interleaved for the still output? and how reliable do you think this solution would be regarding eg. latency? Any thoughts or ideas for how to accomplish this or am I on the wrong course? The images will be fed to a NN Yolo network. Thanks! 🙂
jakaskerl Hi Kristoffer While not directly possible, you can pipe the still images to a ImgManip node and change the FrameType to BGR/RGB and interleaved to False. Might introduce some latency, but likely not too much; depends on the sensor resolution. https://docs.luxonis.com/projects/api/en/latest/components/nodes/image_manip/, you set the properties with ImageManipConfig. Thanks, Jaka
userOfCamera jakaskerl Could you clarify how to do this? When I try, I get this error: [error] Input tensor 'images' (0) exceeds available data range. Data size (614400B), tensor offset (0), size (1228800B) - skipping inference
userOfCamera Solved this by setting frameType as RGB 🙂 resize_rgb.initialConfig.setFrameType(depthai.ImgFrame.Type.RGB888p)
