Dim Target Detection Based on Visual Saliency In Linear CCD Image

To solve the problem that the dim target of line-scan CCD image detection in complex background, a target detection method based on visual significance is proposed. According to the characteristics of line-scan CCD target image, a new feature vector was constructed, double search windows were designed which was similar to the target shape, and the line-scan CCD image´s saliency map was got through calculating the similarity of the center and surrounding pixels neighborhoods´ feature vector in the window. Then the saliency map was clustered and segmented to extract the targets. The experiment results show that the method can effectively suppress noise and background to detect dim targets in the images.