Research And Design Of Intelligent Detection System For Public Bus Emergency

To integrate the development of urban economic, promote the rational allocation of resources and implement strategy innovation-driven development, we have put forward the concept of smart city. Intelligent bus system is an important part of the smart city. It is the best way to solve the traffic problems, but also the artery of economic development and it will be focused by the community.Aiming at the problem, the intelligent vehicle monitoring technology is not perfect and there are few emergencies detection methods. Asynchronous threads ware used to equip the monitor with the ability of "look" and “listen”, between image processing and audio processing module. This paper focuses on the following research:(1) Image processing module: First, the main sports area was grouped according to the trajectory of the passengers. Second, an improved moving foreground extract method was used to extract moving foreground. Then the algorithm extracted the characteristic points in the moving foreground by Harries operator, and used the optical flow constraint algorithm to establish the motion vector filed for characteristic points. At last, the KPA(kinetic potential area) model was built to recognize the emergency. Theoretical analysis and experiments show that, in testing the emergency under different environment, the algorithm in this paper has a success rate of more than 83.9%.(2) A sample discussion and simulation on audio processing module: First, preprocessed of the audio file to obtain audio frames. Then, detected endpoints of audio frames by short-time average amplitude and threshold rate, removed the ineligible instantaneous interference. Finally extracted characteristic parameters of sound, detected abnormal sound by HMM models. Obtained through experiments, combination of MFCC and its differential coefficient have the highest accuracy to detect the audio up to 89%.(3) Designed the smart bus emergency system. It contains the framework of the whole intelligent detection system, functions and relationships among various frameworks, and algorithm flow chart.