It is possible to trigger a manual OCR processing by publishing an MQTT topic "do_ocr" with the payload "1"

On the camera configuration page, it is possible to find and copy the MQTT topic of the camera.

In this example, publish an MQTT Topic "ConPDS/ocr/16f5f10041a17fc66d1eb0b9eeafb03f3fac2bbf/cameras/1/do_ocr=1", which will trigger ConPDS Tracker doing an OCR processing of the current frame in the video stream.

After some milliseconds, you will receive the result in the payload of "results" MQTT topic, and the payload of "do_ocr" is reset to "0".