Research On The Method Of The Perception Model And Coverage Analysis For Video Sensor

Video sensor is an important sensor which is widely used in public places for monitoring emergencies.It plays a big role in maintaining social stability and security.In recent years,with the developing of the "Green City" project,a lots of cities have increased investment in video sensor.In this case,the development of video sensor has experienced a digital and lasting process into a high-definition and intelligent process.However in this process,numerous problems have exposed gradually.In the cases which use video sensor to detect,some of the video images are too fuzzy to recognize the criminal,some of the crime scene are blocked,and the others are not shooting at the criminals,it greatly reduces the efficiency of the video sensor.The problems above reflect that some existing video sensors,layout are not reasonable,some have blind areas and some are redundant construction.The short of the sufficient knowledge about the video sensor and the lack of the effective analysis theory and tools are the very reason that lead to those problems.How to allocate resources effectively is one of the issues which the geographic information system try to resolve.With the help of GIS,we can study the video sensors' features and their influence coverage,which will provide a great solution for the problem of allocation of video sensor.Then we could improve the availability of the video sensor in cities.In this background,this paper try to study the features,perception coverage,optimization of the video sensor,in order to develop a theory about the video sensor's allocation.The research contents and results are as follows.(1)The perception model for video sensor.In this paper,we analyze the principle of pinhole projection,the imaging characteristics of CCD/CMOS and the perspective projection model,then construct a perception model for video sensor.Based on this model,we study the mapping relationship between the image pixel and the point in three dimensional geographic space,and the method that how to obtain the video sensor's position and parameters.At last,we develop the "Video sensor spatio-temporal label collection system"(2)The coverage of video sensor.In this part,we summarize the common obstacle that block the line of sight and express the obstacles by grids.We study the algorithm of bi-valued-grid scanning in order to calculate the coverage of the video sensor.In the end,the "The coverage of video sensor system" is been developed.(3)The strength and features of video sensor's perception.The study design standard parameters to measure the strength of video sensor's perception.Analyze the different features of video sensor's perception,and how the parameters affect the strength of perception.Then we study how the angle between the aspect of video sensor and the aspect of target,and the changes of video sensor's pose influent the strength and features of video sensor's perception.(4)The deployment optimization for video sensor.The deployment optimization problem is abstracted into a combinatorial optimization problem,then we use the searching algorithm of A*to solve it.Finally according to the coverage rate and the strength of the perception,we find a way to optimize video sensors.