Recursive-learning-based moving object detection in video with dynamic environment
- PDF / 1,055,758 Bytes
- 12 Pages / 439.642 x 666.49 pts Page_size
- 56 Downloads / 201 Views
Recursive-learning-based moving object detection in video with dynamic environment Kalpana Goyal1 · Jyoti Singhai1 Received: 5 January 2020 / Revised: 28 July 2020 / Accepted: 11 August 2020 / © Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract Moving object detection is a fundamental and critical task in video surveillance systems. It is very challenging for complex scenes having slow-moving and paused objects. This paper proposes a moving object detection algorithm which combines Gaussian mixture model with foreground matching. This algorithm is able to detect slow-moving and paused objects very effectively. This algorithm uses adaptive learning rate to deal with different rates of change in background. The performance of the proposed algorithm is evaluated on the challenging videos containing strong dynamic background and slow-moving and paused objects using standard performance metrics. Experimental results show that the proposed method achieves 25% average improvement in accuracy compared over existing algorithms. Keywords Background subtraction · Background modeling · Moving object detection · Segmentation · Foreground matching
1 Introduction Analysis and understanding of video sequences is an active research field. The basic step in many computer vision applications like smart video surveillance, traffic monitoring, object tracking, automatic sports video analysis and gesture recognition in human-machine interface etc. is to detect the moving objects or foreground objects in the scene. Moving objects detection has also been used for wide range of applications like activity recognition, airport safety, monitoring of protection along marine border and etc. The moving object detection serves as a pre-processing step to higher-level processes, such as object classification or tracking. Hence, its performance can have huge effect on the performance of higher-level processes. The aim of moving object detection is to extract interesting moving objects in video sequences. These video sequences can have static or dynamic background. Examples of interesting moving objects are walking pedestrians and running vehicles. A frame of a video consists of two source of basic information that can be used for moving object detection and its tracking: visual characteristics or features (such as color, shape, texture) and Kalpana Goyal
[email protected] 1
MANIT Bhopal, Bhopal, India
Multimedia Tools and Applications
its motion information in the consecutive frame. The moving object detection is a task to extract a low-level feature temporally and combines it to the initially segmented object to form a homogenous region in the video segment. The effort is to perceive the movement of pixels at two different instants of time and integrate them to get the relevant information. Once the intensity of pixels belonging to the certain object is sensed at different time condition, the velocity, displacement and other vectors can be computed. Background subtraction is the most common method to iden
Data Loading...
