The super-resolution technology of veterinary B-ultrasound system images is to use the existing low-resolution (LR) veterinary B-ultrasound images to obtain higher resolution (HR) veterinary B-ultrasound image observations by a certain method without changing the veterinary B-ultrasound image detection system. At present, many methods have been proposed to improve the resolution of veterinary B-ultrasound systems. Compared with other methods, the interpolation method is the simplest one. Wavelet transform is a new temporal (spatial) edge domain analysis method that can achieve localization in the temporal (spatial) domain or frequency domain. It has multi-resolution analysis functions and gradually subdivided properties. At the same time, the multi-scale decomposition characteristics of wavelet transform are more in line with human visual mechanism. Some people use wavelets as interpolation basis functions to give applications in information blending and accumulation, and some people use the combination of wavelet decomposition and fractal interpolation to discuss data compression of complex terrain. Artificial neural network (ANN) provides a powerful tool for signal analysis and veterinary B-ultrasound image analysis with its unique self-learning, self-organization and self-adaptive characteristics. Some people abroad have also applied neural networks to image processing and achieved good results, opening up a new way for veterinary B-ultrasound image processing and analysis.

Through some research, some people proposed a super-resolution processing algorithm for veterinary B-ultrasound that combines wavelet transform, B.zier surface interpolation and neural network, and applied this method to the image processing of veterinary B-ultrasound system. The simulation results of these methods in practice show that these algorithms have further improved the quality of veterinary B-ultrasound images compared with traditional methods.