| 摘要 |
In a system for predicting a future crime through an image analysis server including a main control unit including a micro-processor, an image analysis API, an object recognition module, and a transceiver, the image analysis API receives original image data from a digital camera, stores the original image data into a long-term storage memory, chain-codes a portion of the original image data subject to pixel change in the form of a rectangle object, and transmits the square object image enlarged through a digital zoom function to the object recognition module. The transceiver measures transmission traffic together with a center control server of an integral anti-crime center connected through a network and transmits the transmission traffic to the main control unit. The main control unit encrypts the original image data received in the object recognition module to low-quality and low-capacity image data less than the capacity of the traffic measured by the transceiver and compresses the low-quality and low-capacity image data to form a low-capacity image compressed file. Then, the main control unit transmits the low-capacity compressed file to the center control server of the integral anti-crime center through the network. The object recognition module transmits an object image extracted from the rectangle object image and a motion pattern image extracted from a motion pattern of the object image to the main control unit. The main control unit compares a dangerous object and a motion pattern thereof stored in the long-term memory with the rectangle object image, allocates a feature value representing the priority of the dangerous motion, and transmits the encrypted and compressed dangerous event compressive file to the center control server of the integral anti-crime center. |
