SPRUJ28E November 2021 – September 2024 AM68 , AM68A , TDA4AL-Q1 , TDA4VE-Q1 , TDA4VL-Q1
The lens shading module corrects for the lens fall of (loss of light) at the corners of the lens. Figure 6-70 shows a sample image of this phenomenon and the corresponding result after treating with LSC operation.
The image on the left shows a block at the center of the lens and a corresponding block at the edge of the lens. Both blocks are equally illuminated, as such should appear equally bright on the image. However, due to lens fall off, the block at the edge appears significantly darker than the block at the center. The image on the right shows the corresponding result after LSC gains have been applied.
This block (2D-LSC) contains lens shading correction by multiplying an image with a gain factor 2-D map, pixel by pixel. The image is conceived to be in Bayer CFA format having a 2x2 color pattern. The gain factor map is stored in internal MEMORY down-sampled, and is accessed and up-sampled by the LSC module to pixel resolution before being applied to the pixel data. The key difference from previous generation LSC is that the gain map is stored internally in the memory instead of relying on DRAM fetch.
The LSC module does not implement any Region of Interest (ROI) functionality and the LSC gain mask is applied on the entire image. Further, only an 8 bit gain is stored in the mask. The gain can be stored in different formats depending on the range. The 8 bits of gain can be used to represent different ranges based on the lsccfg.GAIN_FORMAT register as per the table below
The offset field is removed from the implementation of the LUT and only the gain field is retained. This is different than previous implementations.