Salient object detection based on HDCT and the prior boundary
Ce Li, Yuqi Wan, Shengze Jia, Hao Liu · 2018
It is difficult to detect the accurate boundary of the object from the frequency information of image. In order to solve this problem, a new salient object detection algorithm which is combined Hypercomplex Discrete Cosine Transform (HDCT) and the prior boundary is proposed. Firstly, we divide the images into image blocks, and the HDCT is performed on each image block. Then the features of the image block are represented by the information of low frequency energy. Finally, the saliency map is calculated according to the prior boundary. The experiments show that the performance of the proposed algorithm is much better than the other classical salient object detection algorithms. At the same time, the boundary of the detected object region is much clearer.