<p>With the increasing number of digital art works, traditional methods of art image classification and recognition are gradually revealing their limitations. In response to this issue, this study aims to explore art image classification and recognition methods based on image segmentation and machine learning, in order to improve classification accuracy and recognition efficiency. A comparative analysis was conducted on key algorithms for artistic image segmentation, and the optimal segmentation algorithm suitable for artistic images was determined. This article constructs a machine learning based art image classification and recognition model, using support vector machine (SVM) as the core algorithm, and designs an image category prediction branch and an image texture feature recognition module to enhance the model’s adaptability to complex art images. The experimental results show that the optimized model exhibits significant improvements in image classification and recognition accuracy for different numbers of nodes, verifying the effectiveness and practicality of machine learning based image processing methods in the field of artistic images.</p>