An Approach for Identifying and Detecting Manholes to Assist Visually Impaired People

According to the National Crime Records Bureau (NCRB), open manholes and ditches in India result in at least two fatalities per day. The likelihood that blind people would trip over the manhole, be hurt, or experience other negative outcomes rose. Therefore, early detection or identification of manholes is crucial to avoiding the challenging situation that blind people may encounter. In this work, an image processing method is applied for real-time implementation using FPGA to recognize manholes and pits. The edge detection approach is used to detect manholes with visible brightness. Since manhole designs are often circular, it is imperative to detect and recognize these shapes for precise estimation. The MATLAB and FPGA platforms are used to accomplish the accuracy comparison.