FEATURE EXTRACTION METHOD FOR UNDERWATER ACOUSTIC IMAGES WITH RADON TRANSFORM

A Radon transform based image feature extraction method was proposed to improve the reliability of feature extraction in underwater acoustic images. In response to the characteristics of low contrast, low brightness, and low clarity in underwater acoustic images, a continuous Radon transform based on two-dimensional functions of underwater acoustic images is proposed from the perspective of Radon transform. Furthermore, a discretized Radon method is proposed, and the calculation process of each order Radon moment and Radon center moment is designed. During the experiment, the underwater pipeline acoustic image was used as the experimental object, and Radon transform was performed on the original image and five transformed images, respectively. The experimental results show that the proposed image feature extraction method based on Radon transform can stably extract Radon moments for images transformed by translation, rotation, scale, and their combination, achieving higher accuracy in feature recognition of underwater acoustic images.