Saliency detection using secondary quantization in DCT domain

Xinyu Shen, Chunyu Lin, Yao Zhao, Hongyun Lin, Meiqin Liu

Abstract: Saliency detection as an image preprocessing has been widely used in many applications such as image segmentation. Since most images stored in DCT domain, we propose an effective saliency detection algorithm, which is mainly based on DCT and secondary quantization. Firstly, the DC coefficient and the first five AC coefficients are used to get the color saliency map. Then, through secondary quantization of a JPEG image, we can obtain the difference of the original image and the quantified image, from which we can get the texture saliency map. Next, considering the center bias theory, the center region is easier to catch people's attention. And then the band-pass filter is used to simulate the behavior that the human visual system detects the salient region. Finally, the final saliency map is generated based on these two maps and two priorities. Experimental results on two datasets show that the proposed method can accurately detect the saliency regions and outperformed existing methods.