摘要
研究了基于Retinex理论的图像增强算法,介绍了图像和视频处理中常用的RGB、HSV、YIQ、YCbCr和Lab 5种颜色空间,给出了均方误差(MSE)、峰值信噪比(PSNR)、信息熵、彩色增强因子(CEF)和结构相似度(SSIM)5个图像质量评价指标的计算公式,并用这5个指标评价了CLAHE算法在5种颜色空间上的增强效果。实验结果表明,相比于其它4种颜色空间,多尺度Retinex算法在HSV颜色空间上取得了最好的增强效果。
This paper studies the image enhancement algorithm based on Retinex theory,introduces five color spaces：RGB, HSV, YIQ,YCbCr and Lab, which are often used in image and video pro- cessing domain, gives the calculation formulas of five image quality evaluation indexes：mean square error （MSE）, peak signal to noise ratio （PSNR）, entropy, color enhancement factor （CEF）, struc- tural similarity （SSIM）, which are used to evaluate the enhancement effect of CLAHE algorithm on five color spaces. The experimental results show.comparied with other four color spaces,multi- scale Retinex algorithm in HSV color space has achieved the best enhancement effect.
出处
《舰船电子对抗》
2017年第1期68-73,共6页
Shipboard Electronic Countermeasure
