摘要
围绕图像识别中的图像去雾模块,在暗通道条件下,对含雾图像的重构算法展开研究。针对去雾图像的颜色失真严重、图像清晰度较低等问题,利用改进的小波函数到暗通道算法多次融合对源图像进行去雾处理,不断调整图像的像素色阶使其达到一个最佳范围,并且使图像的色彩失真在一定程度上获得补偿,以此实现图像去雾的目的。仿真实验表明,利用多次融合的改进小波函数在暗通道的去雾图像重构效果最佳,重构后有更好的清晰度,在多方面的表现均优于其它算法。
Focusing on the image defogging module in image recognition,the reconstruction algorithm of foggy image is studied under the condition of dark channel.Aiming at the problems of serious color distortion and low definition of defogged image,the improved wavelet function and dark channel algorithm are used to defog the source image for many times,and the pixel color level of the image is constantly adjusted to reach an optimal range,and the color distortion of the image is compensated to some extent,so as to achieve the purpose of image defogging.The simulation results show that the improved wavelet function based on multi-fusion has the best effect in the reconstruction of defogged images in dark channels,and the reconstructed images have better definition,which is superior to other algorithms in many aspects.
作者
王义
贺立波
田敏
陈守益
黄祥
WANG Yi;HE Libo;TIAN Min;CHEN Shouyi;HUANG Xiang(Guizhou Branch of China Telecom Co.,LTD.,Guiyang 550001,China)
出处
《微处理机》
2024年第2期34-37,共4页
Microprocessors
关键词
图像重构
暗通道
小波变换
去雾
Image reconstruction
Dark channel
Wavelet transform
Defogging
