Intelligent Moving Objects Detection via Adaptive Frame Differencing Method

The detection of moving objects is a critical first step in video surveillance, but conventional moving objects detection methods are not efficient or effective for certain types of moving objects: slow and fast. This paper presents an intelligent method to detect slow- and fast-moving objects simultaneously. It includes adaptive frame differencing, automatic thresholding, and moving objects localization. The adaptive frame differencing uses different inter-frames for frame differencing, the number depending on variations in the differencing image. The thresholding method uses a modified triangular algorithm to determine the threshold value and reduces most small noises. The moving objects localization uses six cascaded rules and bounding-boxes-based morphological operations to merge broken objects and remove noise objects. The fps value (maximum 72) depends on the speed of the objects. The number of inter-frames is inversely proportional to the speed. The results demonstrate that our method is more efficient than traditional frame differencing and background subtraction methods.