Research on Intelligent Learning Graphics Diagnosis Technology Based on Computer Deep Learning Algorithm

The image feature extraction method and the choice of the classifier are the key factors affecting the accuracy of image classification. The traditional algorithm uses a single image feature and shallow structure to classify images. The algorithm is simple to implement but the accuracy of the result is not high. This paper proposes an adaptive convolutional neural network (CNN) image classification algorithm. By fusing the main color features of the image, CNN is used to extract spatial location features, and for the setting of multi-feature fusion weights, the improved differential evolution algorithm is used to optimize the weights of each feature and improve the classification accuracy of fixed weights. The experimental results show that the classification accuracy of the algorithm is improved by 9.2 percentage points compared with the CNN algorithm, and it has a better classification effect in image classification.