| Summary: | 碩士 === 國立臺灣科技大學 === 電機工程系 === 89 === In our thesis, we construct a surveillance system of combining the technology of sound recognition and image processing. The goal is to detect traffic accidents. The system detects the occurrence of traffic accidents and records the video around the time of traffic accidents. One can use the data to understand how traffic accidents occurred and the duty of traffic accidents. After the pre-processing of the sound, the system recognizes the kinds of the sound by using back-propagation neural network. For saving the image data, we allocate a buffer to save real-time images. When a hit occurs, the system establishes a complete record. Thus, the method will reduce the storage the space.
|
|---|
