| In the process of fatigue driving identification and detection,the selection of sample is the key and difficulty of the whole identification process.In the image acquisition process,often by the outside world more or less affected.And reflection is often the impact of factors.Because the reflective area will have an adverse impact on the performance of the algorithm,if we can reasonably select a reflective sample heavier,then the accuracy of the identification and detection has an important guarantee.Therefore,when designing image recognition algorithm,we need to detect the reflection area of the collected image samples,so as to avoid the effect of the reflection.The system characteristics of the reflective area itself has high brightness and saturation of the relatively low brightness of input processing and analysis procedures,and extraction with associated characteristics of high brightness and saturation area of the picture picture as low brightness processing results.According to the gradient value o f the brightness layer component corresponding to the reflection area edge in the picture,the gradient of the brightness layer of the image is calculated,and the edge of the reflecting area is obtained.Combining two kinds of results to locate the reflecting area.First introduces the related parameters of the reflective region recognition system and needed technology,and then the system is described,analyzed reflective recognition system,based on the demand analysis,the outline design of the system is given,the system module,according to the detailed design of modules of the system are described,combined with the code module results are given.According to the experimental results.Show that the reflective area recognition method of the program system has high positioning accuracy,less reflective area undetected and program process visualization advantages.To reduce the false detection rate of fatigue driving detection system can play an effective help. |
PDF Full Text Request