Study On Local Occlusion Object Recognition Based On Integral Imaging Technology

Vision is the most direct way for humans to perceive the real world.To objectively reestablish the world that they have seen is one of the aims for people.However,the traditional two-dimensional display technology can only simply reflect the content of the spatial scene,lacking the key information such as the depth,to reflect the spatial position of the objects,which cannot meet the needs of people.The Three-dimensional display technology,which can display stereoscopic images with depth came into being.Integral imaging is a promising technology for displaying three-dimensional images with full parallax and continuous viewpoints.It has the advantages such as working with incoherent light and being observed without special glasses.It can be used in many fields such as entertainment,business,medicine,military,etc.However,due to the small aperture size of micro-lens array and the cameras with limited resolution,integral imaging has the disadvantages of small field angle,limited depth of field and low resolution,which needs to be improved.What's more,if some regions of the interesting target object are occluded by the occlusions,on the one hand,the full picture of the target cannot be viewed,one the other hand,the focused plane images of the target objects will be covered with some defocused occlusion information when we reconstruct the captured elemental images computationally,which makes the reconstructed image poor.In order to overcome the problem of poor image quality caused by the low resolution of traditional integral imaging reconstruction images,this paper proposes an image enhancement method based on fast premapping and interpolation algorithms.At the same time,in order to decrease the impact of the obstruction on the target,an image segmentation algorithm and the stereo matching algorithm in the synthetic aperture integral imaging to remove the obstruction,and an method based on the depth map and multi-view images to repair the image are also proposed in this paper.The specific works are as follows:1.The first chapter mainly introduces the research background and research significance of integral imaging technology.First,the history of three-dimensional display technology is introduced briefly.Second,the history of integral imaging technology and the current research situation of the removal of obstruction are described.Finally,the research background and significance of this paper are expounded.2.The second chapter is an introduction to the theoretical part.Based on the principles of integral imaging,firstly,the advantages and disadvantages of several typical image recording systems for light-field information are summarized.Secondly,for integral imaging reconstruction,two methods for three-dimensional reconstruction are introduced and two computational reconstruction methods are experimentally verified and simply compared.Finally,the three performance parameters characterizing the integral imaging 3D display system are briefly introduced,and the mutual constraints between the three parameters and influencing factors are analyzed.3.In chapter three,for the demand of high-quality 3D visualization of integral imaging,a method to improve the image quality based on fast pre-mapping and interpolation algorithm is proposed.This method first uses a depth map to synthesize more viewpoints for a limited number of element images,increasing the amount of scene information collected.Then,the combination of a fast pre-mapping method and interpolation algorithm to reconstruct element images makes full use of the information of the element images and improves the unevenness of the reconstructed image caused by the empty pixels on the reconstructed plane.Simulation results show that the proposed method can make the image more clearer,the brightness and sharpness are better than images reconstructed by traditional method,the focused image is smoother than the conventional reconstruction method.4 In chapter four,aiming at overcoming the problem that the focused plane images of the target objects will be covered with some defocused occlusion information,this paper proposes a occlusion removal method based on image segmentation and stereo matching algorithm in synthetic aperture integral imaging.and an image repairing technology based on depth map and multi-views images to restore the missing target information of the elemental images after the occlusion is removed.This method first uses synthetic aperture refocusing and SMD2 merit function to estimate the disparity range of the scene.Then,the image segmentation algorithm and stereo matching algorithm are used in elemental images to obtain the parallax map.Set a reasonable threshold to classify and label the occlusion.Finally,restore the missing information of the target object in the element image using a image repairing technology based on depth map and multi-views images and perform 3D reconstruction.The simulation results show that the method proposed in this paper can effectively remove the obstruction and repair the target.Compared with the non-occlusion removal process and the non-repair processing method,the PSNR of the reconstructed image of this method is improved by 4.7dB and 3dB,respectively.