Image Aesthetics and Visual Communication Optimization Based on Deep Convolutional Neural Network

With the rapid growth of digital images and visual content, image aesthetics and visual communication face the challenge of how to improve image quality and attractiveness. This paper aims to optimize image aesthetics and visual communication effects using deep convolutional neural networks (DCNNs). First, a large-scale image dataset covering a variety of styles and themes is established and annotated in detail to build an image aesthetic evaluation model. A deep convolutional neural network is used to extract features and aesthetically score images, and the accuracy of the model is improved by combining transfer learning methods. Then, the images are enhanced by an optimization algorithm to improve their aesthetic quality and communication effects. The optimized images have an average improvement of 15% in aesthetic scores and 20% in visual communication effects, which verifies the effectiveness of deep convolutional neural networks in image aesthetics and visual communication. The optimization method based on DCNN improves the aesthetic quality of images and provides more scientific support for visual communication.