Intelligent CCTV-Based Motorbike Theft Detection System in Bike Parks with 3D Slot Identification and Posture Monitoring

Due to the increase in theft and unauthorized access, the security of motorcycle parking facilities has become a critical concern. Traditional security systems often fail to provide complete monitoring and control solutions. This study introduces an integrated approach that reflects two main aspects witch is precise parking slot bounds identification and user activity monitoring within each parking slot to identify threats. The parking slot detection model uses object recognition techniques to accurately identify specific parking slot. Then further process it to get exact slot area as a 3D slot, identifying parking slot bounding box accurately. Meanwhile, the motorcycle security system uses MediaPipe Pose analysis to identify any behavior that could indicate attempted theft by analyzing user activity, identifying suspicious behaviors that may indicate theft or unauthorized entry. There is another model to detect Motor bikes. All these models together develop complex systems that improve the efficiency and safety of motorcycle parking areas.