Research on Image Art Creation System of Integrated 3D Computer Image Vision Digital Media

The accuracy and real-time performance of existing image labeling algorithms are not high. This project intends to study the video sequence feature labeling algorithm for digital multimedia images. A digital multimedia image segmentation method based on wavelet transform is proposed. Then the system adopts the color video segmentation method based on edge detection. Based on visual salience features, the model combines color features and gradient features to describe local image quality. The method fuses the brightness channels of the reference image and the corresponding distorted image, and calculates the similarity between the gradient map of the channel and the gradient map of the reference/distorted image. Target detection is realized by matching the target image with the gray information. Then the color feature and gradient feature are combined organically by means of the mean offset weight method, and a better image quality evaluation effect is obtained. The performance of many evaluation algorithms and has some advantages compared with the existing evaluation methods.