An Improved Shadow Detection Algorithm Based on Multi-features

Aiming at the problem of low accuracy of traditional shadow feature-based detection methods, an automatic shadow detection method fusing multiple features is proposed, which combines a multi-scale pyramid, Mean Shift segmentation, and SVM classifier. Firstly, a multi-scale image pyramid is constructed and the images are preprocessed with weighted median filtering to reduce noise and improve image quality. Then, Mean Shift segmentation is performed on all the images to split them into multiple sub-regions, features are extracted for each sub-region and these features and their corresponding labels are put into an SVM classifier for training to obtain the final shaded regions. The method incorporates an image pyramid mechanism, which reduces the image resolution through downsampling operations, decomposes the original image into sub-images at multiple scales while retaining the main information of the image, and improves the accuracy and robustness of the algorithm by increasing the accuracy and robustness of feature extraction. The experimental results show that the proposed shadow detection method obtains good performance and has a wide range of applications.