P‐8.9: Improved Application of Vector Quantization Algorithm in Demura Data Compression
Pan Pan Yang, Xinglin Guo, Lin Chen, Xiaohong Zhang, Xiaoping Tan, Mingwei Gei, Xiujian Zhu · SID Symposium Digest of Technical Papers · 2024
As an effective compression technology, vector quantization (VQ) has the outstanding advantages of large compression ratio and simple decoding algorithm, so it has become one of the important technologies of image compression coding. OLED, as a light emitting device, is prone to uneven light and dark phenomenon, that is, mura phenomenon, because of its TFT electrical characteristics. The Demura compensation data is written into SRAM to compensate the pixel driving process of OLED display panel to removing mura. Because the Multi‐Photo of gray need to take, the depth of compensation data generated by each gray level is large, so the total bit of compensation data is huge, data compression is required to meet SRAM requirements. This paper analyzes the advantages and disadvantages of VQ‐LBG quantization algorithm, applies LBG algorithm to demura data compression, and improves the accuracy of code vector by optimizing the "splitting" process of LBG algorithm, and optimizes the coding table after the "splitting" to improve compression rate of the algorithm.