Motorcycle image tracking and edge detections based on Simulink software

Abstract

A visual driver support system was developed to reduce accidents involving motorcyclist. The system used Matlab software as a platform to detect motorcycle image. The detection system was designed to detect still images and moving objects images for different resolutions. A motorcycle was defined as the target object in this case. The results showed that the visual driver support system is able to detect image of motorcycle in still and in moving condition. The percentage of correct detection of motorcycle image is 83.3% and 50% for low and high image resolutions respectively.

Real-Time Motorbike Detection: AI on the Edge Perspective

“…Some researchers resort to custom datasets, which unfortunately are not publicly available. These include datasets used by [4,13,14,[32][33][34][35][36][37]. The public MB7500 dataset, published by the authors of [38], stands out for its coverage of occlusion and representative surveillance viewpoints, making it a valuable resource for two-wheeled detection datasets.…”

Detection of Motorcycles in Urban Traffic Using Video Analysis: A Review

“…Illumination variances can be removed by edge detection and binary edges can provide normalized input for feature descriptors. So, the Sobel operator has been used [33] to design collision avoidance systems, detecting the rear view of motorcycles, by defining i nitial hypotheses of possible objects from shadows, wheels and vertical edges [34] and using template matching [35]. However, these approaches are not suitable for CCTV surveillance cameras due to a restricted distance and view angle (motorcycle rear view).…”

“…Some work does not use classifiers e.g. [32] and [35] where a bounding box is used for classification, which is constructed based on width localization of the edge that corresponds to the lower part and upper part of the vehicle. The bounding box size and ratio corresponds to a constant parameter defined a priori.…”

“…The images were recorded during several months (spring, summer, fall) during daytime and under good and medium weather conditions. Specially oriented to motorcycles detection and classification and in some cases tracking, most authors present results working with their own datasets that are seldom made public as in [20], [21], [18], [25], [34], [35], [13], [14], [28], [37], [39], [36], [42], [43], [44], [45], [49] and [53]. This is a significant problem to compare results.…”