Experimental analysis of different studies for using pedestrian detection

This thesis aim is monitoring of pedestrians, pedestrian detection, creation of pedestrian movements following the video with trying different algorithms based on differerence analysis between video frames and experimental analysis of previous studies and comparision of our algorithm and previous studies.This project is made with image processing techniques for detecting pedestrians.This process was done with some different recording videos. The project includes these steps; converting rgb to grayscale , calculating background with grayscale, subtracting video frame to background for each video frame and kmeans algorithm to make threshold in the frame to convert it black and white frame(binary frame), some of regions to delete the noise and smaller than the obvious region on the frames,and doing morphlogic operation for each videos frames.After performing all these steps over frames, the regions with the pedestrians were detected.Moreover,different algorithm is produced and tested for detecting pedestrians. After that, previous studies is done experimental analysis and comparison process with our algorithm.