Based On Opencv Digital Image Stabilization System

Machine vision technology in information science is a hot research field; it acquired through the front of the camera video analysis, classification, identification and a series of intelligent processing, and thus makes judgments on the external environment. Machine vision technology because of its non-interference to the external environment has been widely applied in aerospace, military, transportation and other fields. Front-end machine vision camera system plays an important role in a machine vision system and determines the stability of the whole system performance. However, in the actual work environment, the front of the camera carrier is often difficult to avoid the unpredictable random shaking. This shaking can cause the camera to shoot the video there is no random shaking, resulting in human visual blur. This non-ideal situation would follow-up observation and video processing has caused great difficulties, such as would cause rate increases, the need to occupy more storage space. To eliminate this vector due to camera shake caused by random interference image stabilization technology is used front-end processing machine vision system. Currently, image stabilization can be divided into three categories: mechanical image stabilization, optical image stabilization and digital image stabilization. The first two techniques mainly by the mechanical system and optical system structure, thus costly, difficult integration. Digital image stabilization (also as electronic image stabilization) is a digital video image processing algorithm to stabilize a technique, which can be easily ported to different platforms, easy integration, can significantly reduce the cost of machine vision systems, which increasingly more attention by researchers.In this thesis, digital image stabilization system is studied. First of all, the existing digital image stabilization algorithm is studied and summarized. Summarizes the common characteristics and scope of the algorithm, and made experiments. Secondly, we developed digital image stabilization system based on OpenCV, focus on the digital image stabilization Modeling and implementation of key technologies, the use of the OpenCV function of the secondary development, digital image stabilization algorithm to achieve a high degree of packaging for the future digital image stabilization technology research as the foundation. The experiment showed that the instability of the video for different scenarios, this paper studied the digital image stabilization system can effectively eliminate camera shake on the random shooting of the video.