Research on Pattern Recognition Algorithm Based on Visual Principle

From the perspective of bionics, this paper introduces a pattern recognition algorithm based on visual theory, which regards the data set as an image and realizes pattern recognition by using the perception and recognition method of visual image. Experiments show that the new algorithm based on visual theory not only maintains good recognition effect for all kinds of data, but also is insensitive to noise data and has good robustness. Combined with the voting strategy, this method is applied to multi-category complex data, and compared with the traditional support vector machine method, the results show that the visual method has obvious improvement in classification accuracy and classification speed, especially suitable for complex classification problems with a large number of categories and difficult classification.