| abstract |
The invention discloses a method for detecting and classifying water surface objects based on visual saliency, which specifically includes the following steps: the first step is to perform dehazing processing on an input image; Segmentation; the third step is to use the visual saliency model based on graph theory to calculate the saliency map of the image; the fourth step, the feature vector of the saliency region is calculated according to the edge histogram descriptor, and the multi-class support vector machine based on decision tree is used. The features are identified; the fifth step is to use the image segmentation algorithm based on wavelet transform to segment the water surface objects; the sixth step is to count the data and output the results. The invention can quickly and accurately locate, identify and segment water surface objects such as ships, algae and garbage, and has important application value in the auxiliary field of river regulation. |