A Saliency Detection Method via Color Contradistinction and Background Similarity
Xufan Zhang, Dianhong Wang · 2019
Image saliency detection helps the computer quickly analyze the surrounding environment, locate the interested objects and extract the salient regions from the background. Conventional image saliency detection algorithms usually have high computational complexity, and the detection results seems to be less than satisfactory under complex application circumstances. In this paper, a novel image saliency detection method via color contradistinction and background similarity is proposed, which is effective. In our method, the input image is reconstructed according to block-based compressed sensing for reducing the computational complexity. Then, a weighted local contrast principle and a background similarity calculation framework are designed to obtain two different primary saliency maps. Finally, a weighted fusion strategy is used to combine the two saliency maps to get the final result which has the best detection performance. The experimental results show that the proposed method has good detection performance in terms of accuracy and running time.