CAJ | 학술논문

由于低照度彩色图像存在整体亮度低、对比度低、颜色偏暗和信噪比低等特点,所以经典图像增强算法对其增强效果非常有限.提出了一种利用BP神经网络进行彩色图像增强的算法,并将RGB图像转换成HIS图像,以保证增强处理不引起图像的色彩失真.实验证明:该方法显著地改善了低照度彩色图像的视觉效果,提高了图像整体亮度和图像的信噪比,可调节图像的动态范围,能增强图像的对比度和细节,可增加图像信息熵.
유우저조도채색도상존재정체량도저、대비도저、안색편암화신조비저등특점,소이경전도상증강산법대기증강효과비상유한.제출료일충이용BP신경망락진행채색도상증강적산법,병장RGB도상전환성HIS도상,이보증증강처리불인기도상적색채실진.실험증명:해방법현저지개선료저조도채색도상적시각효과,제고료도상정체량도화도상적신조비,가조절도상적동태범위,능증강도상적대비도화세절,가증가도상신식적.
Because of the low illumination color images have the characteristics of low brightness, low contrast,dark colors and low signal noise ratio(SNR), classic image enhancement algorithms are very limited on their enhancement effects. A color image enhancement algorithm based on BP neural network is proposed. In order to prevent color distortion during the image enhancement, the RGB image is transformed to HSI image. Experiments show that the algorithm improves the low-illumination images' visual effect on enhancing the brightness of the image and the image SNR, adjusting the image dynamic range, enhancing the image contrast and detail, and increasing entropy of the image.