Case Study: Real-Time Object Remove and FOV to Monitoring AI Camera Cluster in Smart Building

Smart building security management teams typically use manual or automated systems to check if the lens is contaminated, blurry, out of focus, or has lens drift. However, in actual applications, objects, vehicles or people often legally enter the camera's field of view, causing a large number of automatic false alarms and increasing operating costs. In order to improve the above problems, this study uses deep learning image detection and object removal technology to achieve a successful case of FOV (field of view) requirements in actual customer scenarios. In order to solve the above problems, this research first obtains the original view of the camera when the camera is installed through the http technology, and then uses the deep learning image detection to find the edge of the target (human figure and the vehicle); secondly, we used to quickly remove the target. The object then uses similar compensation technology to restore the picture and save it as the original file; finally, use the edge point technology in the real-time image to find the important displacement points and the displacement amount to send the event every day. Through the application of artificial intelligence technology, it optimizes the safety management and manpower utilization of application places. This technology mainly uses the artificial intelligence analysis of the Field of View (FOV) to find out the location and travel characteristics of the special edge points in the image as the analysis of the different ratios of the two images before and after, so as to ensure that the monitored and desired Whether there is any abnormality on the target, and then ensure the safety of the place and personnel.