IMA-BLC: Iterative Median-Averaged Adaptive Black-Level Correction Method
Yaoyi Chen, Yujie Huang, Mingyu Wang, Wenhong Li, Xiaoyang Zeng · IEEE Transactions on Circuits & Systems II Express Briefs · 2024
To accommodate temperature changes, typical black level correction methods utilize black pixels to obtain the black level value instead of using a fixed corrected value. However, this type of method is susceptible to bad pixels. Although some sorting-based algorithms can correct these bad pixels, the corresponding storage and computation overheads are large. To solve this problem, this brief proposes a new black-level correction method: the Iterative Median-Averaged Adaptive Black-Level Correction Method (IMA-BLC). The proposed method can remove the effect of bad pixels by shifting the filter window and obtaining median values in consecutive intervals with low hardware overhead. The average of all median values is then calculated to provide the statistically significant black level value. We have implemented the IMA-BLC framework on the FPGA platform. The proposed method has the lowest hardware cost and power consumption compared to some common black level correction methods with bad pixel correction capability.