Using Fuzzy Inference System FIS for Identifying Motion in Digital Surveillance Systems

The aim of the current study is to use the FIS Fuzzy Inference System method to determine motion in digital surveillance systems. In the analysis, there are several methods used such as segmenting, sorting, evaluating and showing the results. After capturing an online video stream with an average of 1 frame/sec, the goal of the proposed method is to translate these frames to their corresponding representation of pixels and then use these frames as inputs for the job. The output is evaluated based on these inputs. The idea is to compare the average pixel representation of the current frame boundaries (column vectors) with the corresponding column vectors in the next frame, in order to find out whether there is any motion detected by comparing the average of the calculated column vector for the ith frame with the corresponding column vector in the i+1th frame. This operation leads to extracting 8 averages and it is considered as inputs to the fuzzy inference method. There is one output that will detect whether there is any motion detected or not. By designing a set of rules and then analyzing the results, a comparison of the averages is held.