Part Number: TDA2EVM5777
Other Parts Discussed in Thread: TDA2
Hi experts:
I have a project that using camera module with ISP IC which output format to 720p's YUV422 10-bits.
I created an new usecase (Capture -> Alg_custom_frame_copy(DSP1) -> Display(YUV420sp_uv)) for testing this camera.
I'd like to do color space transformation on Alg_custom_frame_copy node,
so I assigned those following custom settings for this camera after calling ChainsCommon_SingleCam_SetCapturePrms()
1: videoIfWidth -> SYSTEM_VIFW_16BIT
2: pInprms->dataFormat -> SYSTEM_DF_YUV422I_YUYV
3: pOutprms->dataFormat -> SYSTEM_DF_RAW16
and i transform data on Alg_custom_frame_copy by the following code:
Int32 row_idx , col_idx ;
UInt8 *src_gray = (UInt8 *)inPtr[0];
UInt8 *dst_gray = (UInt8 *)outPtr[0];
UInt16 *row_ptr = 0;
UInt16 tmp;
for (row_idx = 0 ; row_idx < 720 ; ++row_idx, src_gray += inPitch[0], dst_gray += outPitch[0])
{
row_ptr = (UInt16 *)src_gray;
for(col_idx = 0 ; col_idx < 1280 ; ++col_idx)
{
tmp = (row_ptr[col_idx] >> 6);
dst_gray[col_idx] = (UInt8 )(tmp & 0xFFU);
}
}
memset( outPtr[1], 0x80, 1280 * 720 / 2); // try the gray channel first, so set UV channel to empty
and i got those following error on the display(see figure):
1. one-fourth camera's FOV.
2. image noisy.
3. pixel-gap on x-direction of image.
so, a few questions below:
1. does my settings/code wrong to transform color space?
2. I try 0 shifting on Alg_custom_frame_copy, I will get dark display. So, I wonder how the bit-ordering of input image. How to get that?(camera's data pin is connected to v1d0~v1d9 by pin-to-pin)
3. Due to there is only one-fourth camera's FOV and pixel-gaping, does VIP module output wrong size of image?
4. Or any good suggestion of my project?
Best / Regard
Wxchen
