Data Fields | |
vx_float32 | mean_pixel [3] |
vx_float32 | scale_val [3] |
vx_int32 | pad_pixel [4] |
vx_int32 | tidl_8bit_16bit_flag |
vx_int32 | ip_rgb_or_yuv |
vx_int32 | color_conv_flag |
vx_uint32 | clear_count |
vx_int32 | skip_flag |
vx_float32 tivxImgPreProcParams::mean_pixel[3] |
Mean pixel for each output plane which needs to be subtracted
vx_float32 tivxImgPreProcParams::scale_val[3] |
Scale value for each output plane which needs to be multiplied
vx_int32 tivxImgPreProcParams::pad_pixel[4] |
Pad value for output planes for each side of image
vx_int32 tivxImgPreProcParams::tidl_8bit_16bit_flag |
To enable 8 bit or 16 bit input/output tensor for tidl. 0 means 8 bit, 1 means 16 bit
vx_int32 tivxImgPreProcParams::ip_rgb_or_yuv |
Falg to indicate input is RGB or YUV. 0 means RGB otherwise YUV
vx_int32 tivxImgPreProcParams::color_conv_flag |
color conversion flag for seperating the plane as peer the model requirement
vx_uint32 tivxImgPreProcParams::clear_count |
Number of times to clear the output buffer
vx_int32 tivxImgPreProcParams::skip_flag |
Skip processing