De-rained Algorithm Of Image And Its Application In Lane Line Detection

This paper proposes a single-image rain and fog removal algorithm based on sparse representation and low-rank representation to repair the image to improve image contrast,increase color saturation,improve signal-to-noise ratio and signal entropy,so that the output image can be close to the clarity of sunny days to correct the image color cast to achieve the needs of target detection,target recognition,image segmentation and edge detection,applicable to traffic monitoring,traffic safety and other fields.In this paper,morphological component analysis(MCA)is used to separate the different morphologies of the features contained in the images.The data show that MCA is useful for decomposing images into textured and sheeted smooth parts or for graphic recovery application.Whereas the classical algorithm uses image de-rain with the specific process that a bilateral filter divides the image into high-frequency as well as low-frequency components,the high-frequency component of the contains background information as well as noise,so that image HOG features can be extracted and encoded with a sparse matrix(sparse coding)as well as dictionary learning can be used to obtain high-frequency image dictionaries.To classify the high-frequency dictionary,as mentioned before,the method used here is the K-Means nearest neighbor algorithm,and after getting the classification results in a geometric dictionary as well as a dictionary of rain components,recovering the geometric dictionary as an image and subsequently summing it with the previously stored low-frequency components.You can get the target image...