Hi everyone,
I have a technical question regarding interlaced video capture and conversion.
I am currently working with the SG2000 platform and have written a driver for the TP-9963 chip (which is a CVBS-to-MIPI-CSI converter). The chip receives a PAL analog input and outputs raw interlaced video over MIPI-CSI to the SG2000.
I am able to successfully receive the video stream, but the output image appears half-sized / split (i.e., it is clearly interlaced with no deinterlacing applied). I am only seeing one field (either even or odd) at a time, so the vertical resolution is effectively halved.
I have already configured the (middleware/v2/sample/common/sample_common_sensor.c) scan_mode to VI_SCAN_INTERLACED in the sensor/driver settings, but the issue persists — the image still looks interlaced (half picture)
My questions are:
-
Is it possible to convert an interlaced MIPI-CSI input to progressive output within the SG2000 ISP/VIN pipeline?
-
If so, what is the correct method or register configuration to enable deinterlacing?
I would really appreciate any suggestions, example code, or reference to relevant documentation.
Thanks in advance for your help!