Digital image processing technology has an increasing influence on the art field, especially image segmentation technology. For segmentation accuracy problems in the art of image recognition, an art image automatic segmentation scheme based on edge detection of Canny operator and GrabCut algorithm is proposed in this paper. First, the image will be preprocessed by the RGB model to achieve smooth and sharp image. Then, the edge detection of the canny operator is used to smooth the filtered image and calculate the gradient value. Finally, the GrabCut algorithm is used to minimize the energy function calculation in the RGB color space, and the segmentation region is combined with other regions according to a random method to construct a segmentation map. The experimental comparison results show that compared with the existing segmentation algorithms, the proposed scheme shows better segmentation effect in direct visual and quantitative analysis.