直方图拟合和亮度判决自适应图像增强算法

摘要

The laser active imaging can overcome the disadvantages of passive imaging under low light conditions,but the SNR of the images is usually severely disturbed and enhancement algorithm is needed to improve the con-trast.Because the laser images gotten at different time have great difference,the characteristics of images under night and twilight conditions are discussed.Through analyzing and summarizing its characteristics and classification method, adaptive enhancement algorithm of laser image based on histogram fitting and brightness decision is pro-posed.Quadratic function fitting histogram is adopted to determine image capture time;the best homomorphic enhance-ment threshold is found through the luminosity function,and the images at different time are classified.The experimen-tal results show that the histogram fitting quadratic function has high accuracy,and adaptive homomorphic enhance-ment algorithm can effectively improve SNR and contrast of the images.%激光主动成像能够克服被动成像的缺点在低光照的条件下对目标成像，但是其图像通常受到严重干扰，需要采用增强算法增加图像的对比度和信噪比。不同时间段拍摄的激光图像存在很大差异，文中对夜间和晨昏成像的不同特点进行了探讨，分析归纳了其归类特点和方法，提出了一种基于直方图拟合和亮度判决的自适应激光图像同态增强算法，采用二次函数拟合直方图判定图像拍摄时间，通过亮度函数寻找最佳同态增强阈值，对不同时段拍摄的图像分类处理。实验结果表明，二次函数拟合直方图的判决方式准确率高，自适应同态增强算法能有效提高图像的信噪比，提高图像的对比度。