Research On Image Quality Improvement Algorithms Of Underpassing Interchanges Monitoring System

As an important information gathering link of smart city,video monitoring system can provide real-time,intuitive and omni-directional data information,which has been radiated to every link and field of production and life.As an important part of urban transportation,underpassing interchanges have always been a key area for monitoring systems in the construction of smart cities and safe communities.Therefore,the municipal administrative department lays a large number of monitoring equipment in the underpassing interchanges to achieve a omnibearing and three-dimensional monitoring of the underpassing interchanges.Due to the nonuniform illumination of the location,bad weather such as fog,dust,and rainy days,as well as the accuracy of the acquisition system and transmission circuit,noise is inevitably introduced into the underpassing interchanges monitoring system that made the collected video graphics some Issues,which low contrast,the loss of details such as edges,that affect the efficiency of the use of real and original monitoring information.Therefore,it is necessary to study the image quality improvement algorithms which get the utmost of the performance of the underpassing interchanges video graphics and improve the efficiency of the underpassing interchanges video graphics,it is very necessary to study the image quality improvement algorithms of underpassing interchanges monitoring system.In view of the above problems,the main work of this thesis is as follows:(1)The classical image enhancement algorithm and Multi-Scale Retinex algorithm with Color Restoration are analyzed,Combined with the characteristics of the underpassing interchanges monitoring image,an improved Multi-Scale Retinex image enhancement algorithm of underpassing interchanges with Color Restoration is proposed.A Single-Scale Retinex algorithm is adopted to obtain more detailed information,which then goes through a Multi-Scale Retinex algorithm to balance the information between detailed information and the global information,thus achieving the goal of obtaining more detailed information without affecting the global information.(2)An image denoising algorithm based on three-dimensional Block Matching is proposed by analyzing and modeling the characteristics of various noises.The idea of Non-Local block matching was used,It looks for the similar blocks in the image firstly,then uses the hard threshold linear transformation to reduce the complexity of the Euclidean distance,and finally uses the joint filter to reduce the noise contained in the similar blocks through the similar block domain transformation,and carry on the weighted processing to the similar blocks in the aggregation to get the target block after noise reduction.(3)The combined algorithm of image denoising algorithm based on Block Matching and 3D filtering and improved Multi-Scale Retinex with Color Restoration image enhancement algorithm is to designed which work as a joint filter to alleviate noise.Compared with the classical image denoising and enhancement algorithm,the experimental results show that the image processed by the algorithm in this thesis has better visual effect and higher objective quality evaluation score,and can effectively improve the image contrast and suppress the influence of noise at the same time.