High dynamic range automotive circuit board image acquisition and reconstruction optimization method

Abstract

The application provides a high dynamic range automobile circuit board image acquisition and reconstruction optimization method, through a polarization imaging technology, polarization radiation maps of automobile circuit boards are acquired at different angles, reflection characteristics of each component are recorded, a polarization degree and a polarization angle of incident light are calculated, automatic identification and processing are carried out on residual glare areas, polarization radiation light intensity suppression of a strong reflection area is realized by calculating Stokes vectors and polarization angles of specific areas, exposure parameters are optimized according to reflection characteristics of different materials, overexposed areas are marked through a stripe coding strategy, a customized stripe mode is generated combined with geometric shapes and material characteristics of automobile circuit boards, 2D, 3D and polarization images are fused by using a principal component analysis (PCA) method, a high-quality fusion image without saturated pixels is generated, problems such as glare and overexposure existing in the automobile circuit board image acquisition process are effectively solved, and the dynamic range and definition of the image are improved.

Images