Research Of Auto Assistant Safty System Based On Machinery Vision Technique

Along with the development of the car industry and the transportation, safety problem has become a big problem that shouldn't be neglected. According to the state of the transportation, the traffic accident caused by tired drive nearly amount to the whole of 30%, so many research organizations and autocar companies have developed researches to the detection of tired drive & alarm ways.At present there are many technology researches on the drive's weary condition, mainly on the three directions: Based on drive individual characteristic monitor method, based on vehicles parameter monitor method and based on drive physiology parameter survey monitor method. Because the first two methods must receive pilot individual characteristic or condition then there is vehicles type limit, both of them are not suitablly to be used as monitor system, therefore this article has used to the pilot non-contact type observation of eye condition monitor method, which calculates pilot's weary degree with one kind of improvement PERCLOS algorithm is extremely effective and has the practical significance. Studied has been done in the current digital video frequency picture technological development foundation, this article proposed one kindof low cost high performance digital video frequency picture gain and the processing system design: Take the high speed DSP digital signal processor as the core processor, the gathering processing image data, and has conducted the research in this system foundation to digital picture solution compression algorithm MPEG, in order to gain this article needs digital video frequency image data.The whole system is composed by hardware electrocircuit and image disposal process in car assistant safety system; TMS320C6711 from TI Company is used in the design of hardware circuit and CH374 is chosen as USB video imput meet, which include digital camer's meet circuit and image storage circuit. The next step is system software part, which include some progress of machinery vedio dispose software arith on vedio frequency decode arithmetic,the determinant of the condition of the drive's weary,the way of ash degree projection orientation and distill of eye's character.The research and exploration very much, enormous , remaining to be gone a step further owing to that machine optesthesia technology image handles the systematic application potential brought forward by thesis. Have been in progress finally in the thesis to studying summing up and looking into the distance , have brought forward future research direction.