Research on Visual Improvement of Image Aesthetics Based on Multi-Feature Joint Learning

Visual improvement of image aesthetics plays an important role in image research, but there is a problem that the improvement effect is not ideal. The homomorphic filtering method cannot improve the aesthetic visual effect of the image, and the feature extraction time is slow. Therefore, this paper proposes a multi-feature joint learning method to study the visual improvement of image aesthetics. Firstly, gradient operators and Gabor transforms are used to train the texture features and edge features of the image to reduce the interference factors in the process of image feature extraction. Then, the multi-feature joint learning method is used to analyze the visual improvement of image aesthetics, and the visual improvement effect of image aesthetics is comprehensively analyzed. The simulation results show that the proposed method is better than the homomorphic filtering method in terms of the accuracy and extraction time of image feature extraction, which can effectively improve the aesthetic visual effect of the image.