Lepton camera image overlap

I’m using openmv to get thermal image through flir lepton 3.5camera lense. I’ve tested the lense on flir’s software and everything seems fine.
However, when i use it on the openmv board. Sometimes the image from the previous frame would pose a layer on the sensor and cause all the later images captured by the camera to become useless.
Here are the example of the overlap I’m talking about.

I need to leave the setup in a remote experiment location. I will probably go to the site once a week to switch out the sd card and check the data. But if the image don’t work properly during the collecting phase, there is no way for me to tell. And the whole dataset will be compromised. Is there anyway I can fix this issue?

Hi, I’ve only seen that when your shutter on your lepton sensor is broken. When the sensor does flat field correction FFC it closes the shutter, calculates the default temperature array the sensor sees, and then opens the shutter and subtracts that from the current image.

Please check that the shutter is closing when the sensor starts FFC.