CMOS图像实时增强预处理研究及实现

Study on real-time image enhancement pre-processing for CMOS sensor

针对CMOS图像传感器固有的不足,本文提出了一种图像增强预处理的算法及实现方法,首先介绍了预处理总体框架,结合实时性、可移植性以及硬件性能的要求,着重讨论图像增强预处理主要功能模块的算法实现,包括伽马校正、色彩插值、色彩校正、色彩空间变化、图像增强、增益控制、对比度调整以及自动曝光等。实验结果表明:该方法有效地改善了图像传感器图像的质量。

In order to compensate for the inherent limitation of CMOS image sensor, an algorithm of real-time image enhancement pre-processing and its implementation are presented. The framework of the pre-processing algorithm is introduced firstly. Considering the real time, transplantation characteristics and the hardware performance required by the method, the major function blocks of the pre-processing algorithm are discussed in detail, including gamma correction, color interpolation, color correction, color space conversion, image enhancement, digital gain control, auto contrast adjustment, and auto exposure. Experiment results prove that the proposed method effectively improves the image quality captured by CMOS image sensor.