Monitoring Target Detection And Target Classification Exercise Cut - Video Surveillance, Video -

Almost every video-based Monitoring System Target detection is a step from the start

. The so-called target detection, is the target object corresponding to the region's image from video sequences separated. The next track on the target and understanding the behavior of target detection and recognition are in the correct target on the basis, so this is Video Surveillance System is very important and very significant step.

Target detection usually include cutting and target classification exercise two parts.

Cutting exercise Cutting campaign video sequence is a video image processing an important and difficult problem. In the video surveillance system, it is to complete the task is to detect and monitor the scene Traffic Tools, the object corresponding to the region, detected movement group is next target tracking and target the key source of information behavior analysis. Cutting exercise, we can not consider only a simple change in the pixel image corresponds to pixels of moving objects changes, weather conditions, lighting changes, shadows move and the chaos caused by repeated movement and so the pixel will change to the campaign cut with high efficiency and reliability to affect. Currently, most sports are cut in the space of image information to be completed. Sports commonly used methods of cutting

Traditional background subtraction method Traditional background subtraction method is commonly used method of cutting movement, especially those with a relatively static background of the circumstances under way. Them through the current difference between the image and the reference background model to detect differences between the pixel motion region. Various processing methods based on the background model and background model in the process of updating the different methods and different classification. The simplest background model Ah, that is, a temporary mean image, the current scene is a still image similar to the same background. The limitations of this approach is that it has nothing to do because the light changes or the occurrence of the incident scene due to the sensitivity of dynamic change much. To reduce the dynamic changes in scene cut to the impact of movement, most researchers focus more on how to build adaptive to changes in external conditions as the background model is updated.

Complex background subtraction method Complex background subtraction can be divided into two types of non-recursive and recursive. Non-recursive Technology Is used to adjust the window method to assess the background. Video frame on the first cache, the cache based on changes in each pixel to evaluate the background image. Non-recursive techniques from those who have not cached in the frame of history, has a strong adaptability. On the other hand, experienced a similar slow-moving traffic, because of the need a large enough cache, storage space has obvious demand. Recursion does not need to retain the context of assessing the cache. Instead according to each input frame to recursively update the background model. The results lead to long ago entered an image remains on the impact of the current background model. Compared with non-recursion, recursion requires very little storage space, but in the modeling process, any one of the model error will result in a long time effect. Most Program Index weights are included, use only the background pixels as the update to remove decision-making history is not correct feedback.

Target classification

Different sports scene in the region in monitoring the movement corresponds to different targets. For example, traffic control traffic control scene Camera Captured video image sequence may contain the movement of pedestrians materials, vehicles, birds, and the flowing movement of clouds ... different regions corresponding to different scene monitoring the movement target. For example, road traffic monitoring camera monitoring the scene to capture the video image sequence may contain the movement of pedestrians materials, vehicles, birds, and the flowing clouds and so on. How to correctly target the movement of other moving objects with a distinction, for further tracking and event analysis, which is very important.

by: gaga

POST COMMENT

Home Workout Tips - What You Need to Be Aware Of When it Comes to Home Fitness Programs Factors You Need To Consider When Making Lcd Monitor Deals How To Buy Cheap Lcd Monitors How To Buy Lcd Monitor Don't Leave Laptop Batteries In Direct Sunlight How to recover file after computer crashes? Lcd Tft Monitors: Your Guide To The Best Deal Install This Latest Honda Dvd With Amazing Features Blood glucose monitoring - Simple Diagnostics Gas Monitors Device with utmost importance Unsecured Debt Relief - How Debt Settlement Programs Work Where To Buy Printer Ink Cartridges Memory Foam mattress is growing in popularity