Image Highlight Elimination Method Based on the Combination of YCbCr Spatial Conversion and Pixel Filling

In optical 3D measurement, the highlight caused by specular reflection often affects the measurement result, and causes great measurement errors or the inability to complete 3D reconstruction of measured objects. This paper adopts a highlight elimination method based on color spatial conversion and pixel filling theory. It first converts the collected image from RGB to YCbCr, so that the image brightness and chroma are separated. Then, it normalizes the brightness and uses the pixel filling theory to weaken the brightness of its highlight position. The experimental results show that the method can effectively eliminate the specular reflection on the surface of the object within a certain range.