An Underwater Image Classification Algorithm Based on PCA and D-S Evidence Theory

摘要

In this paper, an integrated underwater image classification method is proposed by combing D-S (Dempster-Shafer) evidence theory and Principal Component Analysis (PCA) method. First, underwater image texture features are extracted and compressed by using PCA method. Then, the correlation coefficient between image feature and classification pattern is calculated, and the fusion belief function of every image classification pattern can be obtained through D-S evidence theory. Finally, by comparing the results between single PCA feature components and fusion data recognition, the superiority of the proposed integrated method in image recognition is demonstrated.