Image Processing And System Research Of Distributed Aperture Imaging Array

Nowadays,infrared imaging technology is indispensable in military,industry,security,medical and many other related fields.Therefore,at home and abroad,infrared imaging system has become a research hotspot in many fields,including target detection,recognition and tracking.Especially in the medical field,tomography technology is often used to determine pathological features such as the size and specific location of the disease body.However,due to the limitation of equipment and the immaturity of existing imaging technology,it is still unable to directly obtain high-resolution images at this stage.Aiming at the urgent needs of the above medical field,in order to improve the image resolution to improve the accuracy of disease diagnosis,this paper studies the distributed aperture imaging system to improve the resolution of images obtained by the imaging system through super-resolution reconstruction technology.Distributed aperture is a kind of structure which imitates the principle of synthetic aperture.A camera architecture simulating synthetic aperture,is built to solve the problem that single camera can not meet the requirements of large field of view and high resolution at the same time.Distributed aperture camera array is a large aperture camera system that combines image information to obtain multiple images from multiple viewing angles.In order to solve the problem of low resolution of infrared images in the medical field,this paper carried out theory and application research on based on distributed aperture imaging system and image processing.The specific research contents are as follows:(1)Studying the synthetic aperture and combining the existing camera array imaging method,after reasonable optical design and industrial design,a distributed array infrared imaging system with simple structure and high real-time performance is built.An infrared imaging system of 2×2 grid camera array is taken as the hardware of this paper,and multiple infrared images of different angles in the same scene can be acquired quickly by the upper computer software.(2)The resolution of infrared image is improved through image algorithm processing.The distributed array imaging system was employed to obtain four images with pixel-level displacement.One image was selected as a reference to register the other three images.The projection of convex set algorithm is used as the image reconstruction algorithm in this paper to obtain a high resolution infrared image,and the dual-scale fusion algorithm is served to fuse infrared reconstruction image and visible image to obtain complementary information of the two images and improve the subjective visual image effect.An improved generative adversarial neural network reconstruction method is applied to obtain infrared super-resolution images.The final experimental results show that a super-resolution reconstruction is realized by the infrared imaging system of distributed array camera,and the image resolution is increased by 8 times,from 400×300 to 3200×2400.Compared with the initial infrared image,the mean value and standard deviation of the reconstructed infrared super-resolution image are improved while the entropy value is basically unchanged.The fusion algorithm of projection of convex set algorithm and the improved generative adversarial neural network in this paper achieves the purpose of super-resolution reconstruction of infrared image,improves the clarity of infrared image,can meet the application requirements of infrared super-resolution image in medical field,and has a great development prospect in the field of public security.