Research And Implementation Of 3D Reconstruction Method Of Occluded Target Based On Synthetic Aperture Imaging

Along with the rapid development of computer technology and applications,people are no longer satisfied with the information display method of two-dimensional plane,and more pursue the three-dimensional model of reconstructing objects.In the field of three-dimensional reconstruction,manual modeling of target objects using modeling software is currently the most commonly used method,but manual modeling requires a lot of manpower and material resources,and its reconstruction effect is often unsatisfactory.The technology of three-dimensional reconstruction of object based on visual information,using digital camera as image collector to measure three-dimensional object in a non-contact way.is not limited by the shape of the object,the reconstruction cost is low,and the reconstruction speed is fast.It can realize full-automatic or semi-automatic modeling.It is an important development direction in the field of three-dimensional reconstruction.However,there are still some problems in the current mainstream vision-based three-dimensional reconstruction methods.When reconstructing scenes containing occlusions,the above methods will lead to the reduction of reconstruction accuracy due to occlusions.In view of the above opportunities and challenges,we use camera array as image acquisition equipment.According to the characteristics of synthetic aperture imaging technology,a three-dimensional reconstruction method based on synthetic aperture imaging for occluded objects is proposed.The main work and research contents of this paper are as follows:Aiming at the problem that the reconstruction accuracy of the existing vision-based three-dimensional reconstruction methods decreases due to occlusion,we uses the characteristics of synthetic aperture imaging to achieve "perspective" occlusion,and proposes a three-dimensional reconstruction method of occluded objects based on improved synthetic aperture imaging,which utilizes multiple viewpoints of the camera array.The effective information of the object reduces and eliminates the influence of occlusion on the three-dimensional reconstruction of the object.The specific implementation steps are as follows:In the first step.the synthetic aperture image focusing on the occluded object is obtained by using the synthetic aperture imaging method,and then the occluded area is labeled and removed.In the second step,the synthetic aperture imaging method is used to focus the occluded object digitally,in which the occluded area labeled in the previous step will no longer participate in the calculation.The third step is to extract the pixels at the corresponding depth and transform them from the image coordinate system to the camera coordinate system and save them to complete the reconstruction.In addition,we have designed and implemented a system which can remove the occlusion and reconstruct the occluded object in three-dimensional.Starting fiom the overall design of the system,the system is divided into three modules:occlusion object marking module,object reconstruction module and front-end view module.Each module is coded and integrated to form the system,which provides some reference for the design and development of the same type of three-dimensional reconstruction system.The experimental results show that the proposed three-dimensional reconstruction method based on synthetic aperture imaging can obtain dense point cloud results.In the scene with occluded objects,the proposed method of three-dimensional reconstruction of occluded objects based on synthetic aperture imaging can significantly reduce and eliminate the influence of occluder on the results of three-dimensional reconstruction of occluded objects.At the same time,the experimental results are qualitatively analyzed and discussed,and the advantages and disadvantages of the proposed method are illustrated.