研究LCoS像素级图像亮度调整的优化方法

杨中东 王鹏 李晓慧 孙长库

（天津大学精密测试技术及仪器国家重点实验室 天津 300072）

摘要：在计算机视觉研究领域，采集到的图像通常由于被测物体局部高亮或局部镜面反射导致局部过饱和，从而影响测量系统的后续数据处理。针对此问题，根据反射式硅基液晶能够逐像素调节入射光的发射率的性质，采用高解析度硅基液晶（Liquid Crystal on Silicon，LCoS）和图像传感器（CCD或CMOS）相结合，设计了基于LCoS的图像亮度调节硬件系统。根据硅基液晶与图像传感器的光学性质，确立了摄像机模型像平面和LCoS掩膜平面之间的映射关系，提出了基于空间映射的像素对应算法。实验证明，该算法准确、可靠，可以实现图像亮度的像素级调整。

关键词：计算机视觉 高解析度硅基液晶 图像亮度调节 空间映射 像素级 像素对应算法

中图分类号：TP242.6 TH74 文献标识：A 国家标准学科分类代码：460.40

Study on Pixel-level image brightness adjustment method based on LCoS(www.xing528.com)

Yang Zhongdong Wang Peng Li Xiaohui Sun Changku

（State Key Laboratory of Precision Measurement Technology and Instruments，Tianjin University Tianjin 300072）

Abstract：In the research field of computer vision，because of the partial high light of the meas-ured object or partial specular reflection，the phenomenon of image partial over-saturation occurs from time to time.This phenomenon may lead to the failure in subsequent data processing that is the core of the measurement system.Aiming at this problem，based on the characteristic that the emis-sion rate of reflective liquid crystal on silcon（LCoS）can be adjusted per pixel，a hardware system of image brightness adjustment was designed based on the combination of high-resolution liquid crys-tal on silicon（LCoS）and image sensor（CCD or CMOS）.According to the optical characteristics of LCoS and image sensor，the mapping relationship between the image plane of camera model and the mask plane of LCoS is established，and the pixel mapping algorithm based on space projection is proposed.Experiment results proved that the proposed algorithm is accurate and reliable，and the image brightness adjustment per pixel can be achieved.

Keywords：computer vision liquid crystal on silicon（LCoS） image brightness adjustment space mapping pixel level pixel mapping algorithm