Intelligent monitoring of home security technology push Chen renovation
Time : 2011/12/20 Hits : 6155 Font : Small Medium Larger [back]
Home security is inseparable from video surveillance systems, video surveillance system has also been widely used in residential security monitoring, fire monitoring, traffic violations, traffic control, military, and banks, shopping malls, subways, airports and other security precautions in public places, and will have a wider range of applications.
However, the existing video surveillance systems typically rely on manual methods to complete the monitoring, alarm, recording a series of processing, not fully real-time proactive monitoring role. More product demand and broader market applications require security products in the future must have the intelligence analysis and processing capabilities. From the current level of technology, the digital image processing technology has become more sophisticated, artificial intelligence is also rapid, intelligent analysis of how the function of the image into intelligence products to various companies and manufacturers have become the focus of attention. Intelligent analysis of intelligence with advanced video surveillance system, not only can greatly enhance the ability to monitor and reduce the hidden dangers, while saving manpower and resources, reducing costs.
Intelligence in monitoring developments in the field
In the early nineties of last century before to simulate the device as the main CCTV system is the earliest first-generation monitoring system, the technology is mature, stable, and in many practical projects have been widely used, in some cases this The technology vision of people's attention are still among the. Truly intelligent monitoring is the development of computers and network with innovative phenomena. The mid-nineties, computer processing capabilities and video technology enables people began to use computers to high-speed data processing capabilities for video capture and processing, the resulting raw digital monitoring system called second-generation video surveillance system . From the late nineties, with the network bandwidth, computer processing power and storage capacity to further improve, and a variety of video processing technologies, digital video surveillance into a new era. Intelligence in Security in more and more important. Thus, the third generation of monitoring system is web-based, digital video compression, transmission, storage and playback as the core, intelligent image analysis for the characteristics of the technological revolution. Intrusion alarm, illegal stay, items left behind, moving target tracking and more. Intelligent analysis of these advanced features into a modern security monitoring products, a major feature.
Intelligent algorithm related
Security can not fail to mention that intelligence, that intelligence must be mentioned algorithm. Algorithm, as behavior analysis and image processing of the soul in future product development and implementation of the program plays a leading role. Monitoring of the area continued to expand, the market continues to develop, make great need of video information processing, intelligent video surveillance systems for the face of such a large amount of information, good technology and structure is a key problem. In the remote monitoring process, in order to reduce the pressure on network transmission, the equipment front end module of some intelligent analysis of the coding module embedded into products which go through the front of the intelligent video analysis to extract useful information only, and its compression and transmission encoding , greatly reduces the amount of data transmission, and image analysis in the decoded part of the software portion of the back-end processing. This division of labor-type thinking than simply relying on a centralized software or hardware to be better and more effective treatment.
H264 codec with other algorithms, the video is no single reason why intelligent algorithm standard, not ASIC technology, it is because of its complexity and diversity of target behaviors, makes it difficult to develop a standard set of rules applies to all industries. Integrated intelligent algorithm is an emerging class of science and technology, many effective start with intelligent analysis algorithms are developed abroad, but also a considerable part of the algorithm is either too complex, not suitable for real-time requirements; or less adaptive , too much dependence on the scene; or are still in the stage of theoretical study, still relatively far away from the product. Truly complete independent intellectual property rights of domestic intelligence surveillance products rarely, the vast majority of technology products are from the United States, Europe, and Israel and other developed countries. Most of the domestic production and installation of video surveillance systems still remain in the area of ??non-intelligence, surveillance system mentioned in the "smart video surveillance", in fact, still in the general concept of network video surveillance.
Most digital or IP monitoring and surveillance, said the senior does not have the intelligent analysis of the content. There are many companies in this country has invested a lot of intelligent human resources research and development, and moving target detection in the traditional has done a lot of improvements achieved breakthrough results. Intelligent products with independent knowledge of security products, to improve the overall domestic intelligence surveillance, raise domestic products in the international competitive position has a very important role.
From the algorithm point of view, monitoring the technical aspects of intelligent products, there are a lot of research to be done, but it also gave us an opportunity, who control the future development of the core technology will be able to dig the first pot of gold, accounting for the relevant area of ??the market. In the intelligent image analysis, moving target detection and tracking high-level intelligence analysis should be inside the most basic functions. Is the intrusion alarm, illegal stay, packet loss and other artifacts based on behavior analysis. Here's a simple analysis to track the process of object recognition.
Specific application examples
Algorithm implementation can be considered as a modeling process that can be understood as a kind of applied mathematics projects. Moving target detection and recognition is the change from the foreground image sequence will be separated from the background, digital image processing is a major part. Changes in weather and light, other objects in complex background interference, movement of the shadow and the camera motion, jitter and so will target the extraction process difficult. Therefore, the right moving target detection and accurate tracking in video surveillance system has become an important technical point. The whole process can be simply divided into the following steps: video pre-processing, object detection and recognition, object classification, object tracking, behavior analysis, and rule creation.
Video pre-
In fact, the video preprocessing is to reduce the noise, outstanding track the target. Two types of image noise, one is the source of noise, a class of observational noise. Changes in ambient light, background, leaves, rain and snow and so on are the source of noise; and observation noise mainly refers to the camera jitter, interference of transmission lines. Objective of tracking the noise extraction process will cause trouble. In terms of eliminating noise, the method commonly used for digital image stabilization, adaptive learning background, thresholding, morphological filtering, binarization and so on. Such treatment can eliminate the most noise in the image, reducing its interference on target detection.
Object Detection and Recognition
Moving object extraction process using foreground moving objects are, you need to target a recognition process. This recognition is divided into single objective and multi-objective artificial identification region recognition. For a single object tracking, it can be selected from the video in a man-made objects of interest, and its track; for the regional multi-object, it can be divided into a region, only the object into the region to identify and tracking.
Target Classification
For the control region is not a single object tracking classification. Even if the object occlusion occurred during exercise, when the object to satisfy linear motion, the background changes relatively smoothly, the algorithm can still accurately track the object. Environment for complex nonlinear motion of objects, we are in the clinic - to single-target tracking based on the target feature to add some information to help identify the track. Target classification generally used in case of multi-objective single-target tracking, multi-objective or multi-target tracking environment. When the environment when there are multiple moving objects can be the target of each movement are extracted, in accordance with the position, velocity, shape, texture and color and other relevant characteristics of their classification. First, the characteristic information of interest to target saved, and then track it. When tracking objects in the video appears and then disappears, we re-use previously stored feature information and the current picture of all moving objects that appear to match. By comparing the results to appear in the object can determine which one is just tracking the target.
Behavior analysis and rules to create
If the object tracking process is a purely technical process, then you should belong to the advanced behavior analysis the scope of the intelligence. Simply put, behavior analysis is man-made rules created based on the automatic video image analysis and extraction of video sources in the process of critical information. Was created because of the rules related to the environment, such as the scene is different, set the conditions for triggering the alarm and regional are also different. Therefore, the development of the intelligent security, the software side is the focus of the work of intelligence analysis as far as possible toward the direction of human development, in a simple setting, you can make intelligent analysis of the entire image has a self-learning process of adaptation, further minimizing of factors, to reduce people's workload.
Summary and Outlook
With the increasingly rapid development of China's security industry, the market expanding, the company engaged in the security industry is also increasing. For the promotion of intelligent security products company, only to seize the pulse of the security market, grasped the key technologies in order to properly respond to future market demand and continuous innovation, constantly finding opportunities in the market competition.
Abroad, the Chinese intelligent security technology is still a long way to go, but it is certain that as the number of frequent domestic and international technology exchange increased, doubling the number of domestic talent and R & D environment, perfect, these gaps will be more to smaller. Hope that all domestic companies in the security industry but also in the protection of intellectual property rights under the premise of the company, a lot of communication and exchanges, and constantly improve product security industry norms and market norms. Win the market in establishing the premise, constant technological innovation, the intelligent image processing technology as the core of the intelligent video surveillance products in various industries and fields expanded, people's lives to create a peaceful and secure environment.
|
