High resolution image has many advantages,such as clear picture,rich color,and so on.However,in the real world,the imaging process is affected by many factors,and it is often unable to obtain high-quality images.In order to solve this problem,we can consider using super-resolution reconstruction technology to recover the details of image loss and improve the resolution.With the development of deep learning,image super-resolution reconstruction technology has made a great breakthrough,which is the focus of domestic and foreign scholars in recent years,by virtue of its unique advantages,convolutional neural network has become the main method to study computer vision.Therefore,this paper uses deep learning technology to build a super-resolution network model,and designs and implements an image definition enhancement system.The main contributions of this paper are as follows:1.Firstly,this paper introduces the research background and basic knowledge of image super-resolution reconstruction,and analyzes the research status of this technology,focusing on the application of deep learning in super-resolution reconstruction.2.based on the idea of generating a confrontation network,a hyper branch network model based on generation of confrontation network is proposed.The objective evaluation indexes of the model and the two existing models are compared.3.Based on the idea of residual network,we propose a model of super division network based on residual network,and compare it with the model based on generated countermeasure network,and choose the model with the best performance as the core model of the system. |