自适应Retinex算法在高动态光照环境下的应用

Adaptive Retinex application in high dynamic range illumination environment

现有的多尺度Retinex(MSR)算法有丢失细节、光晕等影响图像质量的问题。针对这些问题提出了一种新型的自适应Retinex图像增强算法。该算法不需要复杂的参数调节,能实现自适应增强,并能消除光晕效应,对细节的表现更加准确。实验结果显示,对高动态范围(HDR)图像在细节恢复和光照还原上,新算法比MSR算法图像质量提高明显。这一算法被用于户外场面监控,能有效地改进在不同光照环境下图像的可识别度。

Enhancing image quality under high dynamic range illumination has significant contribution to many fields in computer vision. The performance of one popular approach, Multi-Scale Retinex （MSR）, has halo effects on the edges and it is inaccurate in recovering the details. To solve these issues, a novel adaptive Retinex algorithm was proposed, which is able to adaptively improve the quality of the image based on its contents. Experimental results indicate that the new algorithm has significant improvement on these issues. As a result this algorithm is adopted for surveillance applications in High Dynamic Range （HDR） illumination environment and has good enhancement performance.