Detection of Parked Vehicles Using Spatiotemporal Maps

Abstract

This paper presents a video-based approach to detect the presence of parked vehicles in street lanes.Potential applications include detection of illegally and double-parked vehicles in urban scenarios and incident detection on roads. The technique extracts information from low-level feature points (Harris corners) in order to create spatio-temporal maps that describe what is happening in the scene.The method does not rely on any background subtraction or perform any form of object tracking.The system has been evaluated using private and public data sets and has proven to be robust against common difficulties found in CCTV video such as varying illumination, camera vibration, presence of momentary occlusion by other vehicles, and high noise levels.

A Comparison between Moving Object Detection Methods Including a Novel Algorithm Used for Industrial Line Application

“…The results of the detection operation for the used objects when the mentioned method is applied are as shown in table (2), the ratio between the detected pixels and the processed time is calculated and recorded in with the other obtained results in table(2). Very small noisy regions of the third object are appeared in figure (16), these tiny parts are not detected according to the resulted values of using this method is as shown in table (2). The ratio is also has a wide differences for the detected objects (table 2, figure 16), although that the first object is bigger than the second object, but the detected area of object 1 is least than the second object (figure 17), it is clear that the processing time of used this method is less than the first method, but this it cannot be used for all objects.…”

Comparative Analysis and Detection of Street Parked Vehicles using Spatiotemporal Maps and Corner Detection Methods

“…To reduce the problems caused by traffic congestion, Intelligent Transportation Systems (ITS) are being deployed to achieve a more efficient use of existing infrastructures [1]. The problem of tracking objects in different framework has been studied generally due to its applicability to infinite practical situations.…”

Long-Term Stationary Object Detection Based on Spatio-Temporal Change Detection

“…S TATIONARY Object Detection (SOD) has recently experienced extensive research [1] due to its contribution to prevent terrorist attacks by detecting abandoned objects [2] and illegal parked vehicles [3]. SOD aims to detect the objects in the scene that remain stationary after previous motion.…”

“…and edge features [14] is again needed to handle BS errors. Moreover, [3] detects parked vehicles over time using stable keypoints instead of BS. Many SOD challenges addressed in previous research are related to BS difficulties with illumination changes, crowds, intermittent object motion and required temporal adaptation.…”