Design Of The Video Retrieval And Identify System Of Road Diseases

With the development of highway,the pavement maintenance timely appears very important.This makes the technology of road disease detection is also urgent to be enhanced.The traditional processing method based on artificial already can't adapt to the needs of the development of road,because they have the following shortcomings such as slow, dangerous, affect traffic and non-accurate. Therefore, research an automatic identification system of road disease is imperative.This paper designs the road diseases video detecting system with Dalian road resources as research background.By using high rapid and precise cameras to shoot road video quickly,with fast processing of computers,to get distress information and identify disease types.However,because of the uncertainty of shooting environment and distress information's weak-signal,it is not easy to get efficient automatic detection algorithms.This topic needs to solve the key is the road diseases video segmentation algorithm and identify types of diseases. Video lens shear tests using algorithm of brightness-differences combining with lens boundary coefficient.For the gradual shot changes detection,color-differ-ences are used as detected feature.And use the texture characteristics combining with composition theory method to extract key frames.The above video segmentation algorithm is proposed to realize the pavement diseases the accurate positioning.Road disease recognition is using improved histogram concavity analysis method up to road distress image segmentation purpose,The improved algorithm can effectively filter except noise interference. And projecting binary image onto the abscissa and y-axis and get the projection vetors.Finally, using the template matching of pavement disease to identify disease types.This scheme's identify results basically satisfy the need of user.In this paper,Ⅰhave compiled a set of software with OOP method.Using Microsoft Visual C++6.0 as development platform,Ⅰhave checked up the algorithm.The experiments obtained satisfactory results.Video detection system of road disease provides a favorable decision support for making conservation decision, has a strong practical value.