基于LIP的多通道图像边缘检测研究

Study on the edge detection of multi-channel images based on LIP model

由于弱光照图像的成像特点,使得普通的边缘检测的算法难以奏效。20世纪80年代中期由Jourlin和Pinoli提出的对数图像处理LIP模型颇具特色,该模型中的方法对低亮度图像敏感,并且对灰度有界图像的加法运算是封闭的,在边缘检测等方面有着其他方法所不具备的优点。文章基于灰度图像的对数图像处理LIP模型,建立了多通道图像的对数图像处理MLIP模型,并在此模型上进行了对微光图像边缘检测,其效果明显优于经典的Sobel算法。

It is difficult to carry out edge detection of dim images by general algorithms. The logarithmic image processing(LIP) model, presented by Jourlin and Pinoli in the middle of 1980s, is better in the aspect of edge detection than others because the way of the LIP model is sensitive to dim images and the addition operation of the model on gray bounded images is closed. In this paper,based on the LIP model,the multi-channel LIP(MLIP) model is built to deal with multi-channel images and applied to the edge detection of dim images. Obviously, the effect of the MLIP model is better than that of the Sobel algorithm.