Rain streaks removal using total variation and sparse coding based on case based reasoning approach

The impact of rain streaks on single images can make it difficult to recognize the surrounding environment using an outdoor camera. Furthermore, a single image is important to use in numerous areas such as in object recognition and scientific research. Therefore, outdoor images and videos in rainy weather conditions will reduce visibility and damage the performance of computer vision algorithms used for extracting features and information from images. This study proposes a new algorithm as a suggestion for the detection and removal of rain streaks in a single image using total variation and sparse coding to restore images. This proposed algorithm will use a retrieval method from a case-based reasoning approach. The experiments and statistical measurements, namely Mean Square Error (MSE), Peak-Signal Noise Ratio (PSNR), Structural Similarity Index (SSIM), Visual Information Fidelity (VIF) and Blind or Referenceless Image Spatial Quality Evaluator (BRISQUE) are used to distinguish which method has better accuracy. The results demonstrated an advantage for our proposed algorithm for the removal of rain streaks. Medwell Journals, 2017.