Design And Development Of Video Quality Diagnostic System Applied In Intelligent Monitoring

With the development of security monitoring technology,its application field is always extended.At the same time,large-scale video-capture devices put serious challenge to its maintenance services.Nevertheless,there are various video exceptions without any uniform diagnostic criteria and models.Thus,it is a challenged research topic on the using of captured videos to diagnose video-capture devices,the time reduction of maintenance staff to manual identified the cause of device failure,work efficiency improvement.For equipment failure of scene monitoring,we designed and developed a system which based on content of videos to diagnose the video quality.It was designed to diagnose unusual brightness,chromaticity and definition of the video.Additionally,it also able to diagnose the existence of noise,loss of signal,camera vibration based on video data.The system made diagnosis according to the severity level of various exceptions.An integrated video quality diagnostic client was developed.The diagnosis will be stored,thus,the user can access accurate information on the capture device working condition for a certain period of time or a certain spot at any time,therefore.The system can obtain the working state accurately,and is easy to maintain.The main contributions of this thesis are as follows:(1)Designed and implemented a video quality diagnostic system,to detecting and diagnosing ten common video equipment failures in accordance with the image quality,and to providing a friendly and convenient system interface for maintainer,it did help device maintainer to position device failure quickly,thus,it leads them to provide useful decision.(2)Proposed an objective picture quality diagnostic method that takes no reference image.Basing on the characteristics of images,our method does not depend on external reference factors to make the abnormality diagnose of ten common video equipment failures.(3)All-around assessing the detection results of the mainstream edge detector,and proposing image sharpness detection algorithms based on edge detector,in which Canny operator was chosen to detect the definition of video.We has been used a noise detector operator to detect image noise and gained good result.In addition,through the luminance histogram and color histogram to run detection of brightness and chromaticity of the image.The analyses and experiment prove the validity of the algorithm.(4)Module testing for each diagnostic algorithm and the whole system are taken.By testing the system completely with normal and abnormal test image sequences,and compared with the real situation,which had proved the system is good stability and accuracy.