Image Saliency Detection Based on Low-Level Features and Boundary Prior
Chao Jia, Weili Chen, Fanshu Kong · 2020 IEEE 4th Information Technology, Networking, Electronic and Automation Control Conference (ITNEC) · 2020
In this paper, a saliency detection algorithm based on image low-level features and boundary prior is proposed. Firstly, the structure of image patch is constructed. Based on the image patch, the saliency value of patch contrast based on the low-level features is calculated by using image color and texture, and the contrast saliency image is obtained. Second, by using the image boundary prior condition, considering the color difference and spatial connectivity between image patch and boundary area, the saliency value of patch contrast based on the boundary prior is calculated, and the prior saliency image is obtained. Then, the contrast saliency image and the prior saliency image are fused to highlight the salient object and enhance the saliency image, and obtain higher quality saliency image and the precision. Finally, Experiments on international open datasets show that the algorithm is in line with human visual perception and has higher precision and better recall rate compared with the existing mature methods.