Research On Improving The Quality Of Ghost Imaging And Implemention Of Super-resolution

Ghost imaging is a new computational imaging technology which has attracted more and more attention in recent years.Compared with the traditional optical imaging it has some unique advantages.But its imaging resolution is a major constraint to the development of the technology.In order to promote the development of ghost imaging from experimental stage to practical,this paper studies about how to improve the image quality of ghost imaging and realize super-resolution imaging.The work of this paper is based on the image quality of ghost imaging and the algorithm of improving ghost imaging.It analyzes the characteristics of ghost imaging model and seeks an effective way to improve the quality of ghost imaging.The main contents are as follows:This paper explores five classical reconstruction algorithms of compressive sensing,and verifies that it can restore the feasibility of the image under a small amount of measurement conditions.The computational ghost imaging model based on compressive sensing is constructed and the simulation experiment is carried out.The discrete cosine transform matrix and the Gaussian random matrix are chosen as the sparse base and the observation matrix respectively.Reasonable selection of compression sampling ratio and light field free propagation distance,and Fourier transform to establish a mathematical model of light field propagation.So as to establish a complete compression-based computing model ghost imaging system.Due to the noise problem in the reconstructed images of the system,the calculation method of BM3 D algorithm and ghost imaging is proposed.The BM3 D algorithm effectively eliminates the errors caused by the original algorithm by filtering the similarity of the image itself.Set the same experimental parameters to compare the results obtained.The results show that the proposed method can effectively improve the peak signal to noise ratio and visual effect of reconstructed images.Finally,the application scope of BM3 D algorithm in the application of ghost imaging is analyzed.Super resolution reconstruction based on multiple ghost imaging images.The same object from six different angles were collected.The SIFT algorithm is used to register six images,and the information corresponding to the same scene is extracted from the image and matched to the same image,and then the superposition reconstruction is carried out by the method of adaptive interpolation.The results show that most of the images obtained by this method have good visual improvement effect,and the information which is not clear enough can be resolved.However,when the ghost image itself recovers the image,the blurred pixel can not be extracted at the index of the key point,which affects the result of super-resolution reconstruction to a certain extent.Finally,we analyze the advantages and disadvantages of the two methods to improve the quality of imaging,and put forward the improvement suggestions.At last we prospect the development of ghost imaging quality and the development of practical application.