Analysis of Problems Existing in Application of Intelligent Video Analysis Technology

Early video analysis application products appeared on the market, which caused a real uproar, and many special application scenarios and application environments can indeed bring great value to customers. For example, the statistical technology of people flow in shopping malls has brought huge technical support to the data analysis of shopping malls. Such as license plate recognition technology, the value brought to the public security traffic management can not be measured by data.
In the end, the current intelligent analysis is still based on the computer under human control to judge the rule behavior of the target. Many rule behavior computers that have not been defined cannot be judged. For example, in crowd aggregation analysis, intelligent analysis can analyze the internal The number of personnel reached the expected value within a certain period of time, but whether these personnel occasionally pass by or deliberately gather can not be analyzed. This also fully reflects the essence of computer intelligence analysis, which still cannot fundamentally replace human work.
Problems with intelligent video analysis technology
From the point of view of intelligent analysis products, because the quality of these products needs to be improved, the image seems to have obvious delays, jitters, and lack of clarity. Due to the instability of the hardware itself, it is prone to crashes, restarts, and false alarms. However, the limitations of the algorithm of the intelligent video analysis itself can only realize all its functions in an ideal environment, so in practical applications, only simpler functions can be realized initially. At present, intelligent video analysis mainly faces the following problems:
The first is the technical level: factors such as insufficient outdoor lighting at night, severe weather, image compression processing, and limited bandwidth of the network transmission link cause image quality degradation, which brings inherent difficulties to security intelligent video analysis; the target is similar to the background or the background is messy, etc. It leads to difficulties in target segmentation and feature information extraction; it is difficult to model complex abnormal behaviors and events, and the corresponding intelligent analysis algorithm has low recognition performance. The above factors are likely to cause undesirable consequences such as false alarms, missing alarms, and difficult tracking, which restrict the improvement of the actual performance of the intelligent video analysis application system. There are still some problems in the product hardware, including the front-end camera and back-end server. In terms of intelligent analysis based on pixel comparison, the clarity of the image directly affects the result of the comparison, and the clearer the camera, the more pixels there are, and the computational comparison also puts more pressure on the server. For example, the license plate recognition in intelligent analysis, the video resources obtained by cameras with different resolutions, the analysis accuracy rates vary greatly, and the comparison efficiency of the back-end servers will also be significantly different. The analysis capabilities of a single server may be used when using SD It is 20 channels, with an accuracy rate of 90%. When using 2 million HD, the processing capacity may be 5 channels, but the accuracy rate can reach 98%. This problem needs to be effectively promoted in the front-end of the replacement and "cloud computing" technology Can be better resolved.
The second is the industry level: the main confusion is that customers have too high expectations of intelligent video analysis products and the performance of the technology is susceptible to use conditions. Since users have little contact with such products as intelligent analysis, they have questions about the effectiveness of such products. In order to attract customers to exaggerate publicity, some manufacturers did not agree with the publicity effect, which made these customers no longer interested in intelligent video analysis products. The manufacturers have not carefully sorted out the application requirements of most users of the products; the lack of in-depth understanding of the security industry in the product development department has led to the unprecedented function of intelligent video analysis products and lack of pertinence. Furthermore, the lack of an authoritative standard certification system results in users being unable to effectively control the performance of the product.
The third is the market level: At present, many intelligent video analysis products are mostly independent research and development, and these companies need to dilute the early development costs. Therefore, the pricing of such products may be higher than the affordability of ordinary customers, which determines that intelligent analysis products, especially behavior analysis products, can only be used in places with strict and precise monitoring needs, such as prisons, high-end communities, or heavyweight institutions. .
Only basic and common key technical problems can be carried out to create conditions for breaking through the deep application of intelligent video analysis; do a good job in the demonstration of typical projects, and guide users to pay attention to the information provided by intelligent video analysis in the management process; deepen the industry ’s intelligent applications and establish " "R & D + integrator + industry users" cooperation mechanism provides an interactive platform for R & D and application; accelerate the development of standards and specifications, and improve the authority and operability of intelligent video analysis product performance identification, in order to promote intelligent video analysis applications " "Guarantee", intelligent video analysis can only be applied in more industries and fields.
Intelligent video analysis is a new technology that is still evolving and improving. If an enterprise cannot quickly establish its own unique core competitiveness and core products, it is not easy to obtain market voice and continuous development. thing. At present, intelligent video analysis technology is still in the growth stage. The biggest challenge is actually from technology. This is an industry that needs to be able to calm down and concentrate on product and value research. As a technology supplier at the bottom of the industry chain, it still needs to rely on products . As an inevitable direction for the future development of video surveillance, although there are still some problems with intelligentization, with the development of computer technology and intelligent technology, these problems will be well solved, and intelligent video technology will be more widely used.

Shaded Pole Ac Motor

Shaded Pole Ac Motor,Ac Shaded Pole Motors,Shaded Pole Motor,Shaded Pole Induction Motor

Changzhou Sherry International Trading Co., Ltd. ,