Fpga Implementation of Moving Object and Face Detection Using Adaptive Threshold

The real time moving object and face detections are used for various security applications. In this paper, we propose FPGA implementation of moving object and face detection with adaptive threshold. The input images are passed through Gaussian filter. The 2D-DWT is applied on Gaussian filter output and considered only LL band for further processing to detect object/face. The modified background subtraction technique is applied on LL bands of input images. The adaptive threshold is computed using LL-band of reference image and object is detected through modified background subtraction. The detected object is passed through Gaussian filter to get final good quality object. The face detection is also identified using matching unit along with object detection unit. The reference image is replaced by face database images in the face detection. It is observed that the performance parameters such as TSR, FRR, FAR and hardware related results are improved compared to existing techniques.