A trusted waterfall framework based moving object detection using FACO-MKFCM techniques

doi:10.1007/s11042-019-07787-8

2019-06-10

PRE

T. V. Mahalingam *

M. Subramoniam

DOI：10.1007/s11042-019-07787-8

原文链接

原文求助

摘要

En 中文

Object detection and tracking are necessary and challenging activities in several pc visual scene applications such as security, motor vehicle navigating and independent robotic navigating. Video scenery security in an energetic atmosphere, particularly for people and motor vehicles, is just one of the present difficult research study subjects in computer system vision. It is a vital modern technology to attack in opposition to violence, criminal offense and publicised security as well as for effective administration of heavy traffic. In this area, monitoring of motionless forefront objects is just one of the most essential needs for security systems based upon the monitoring of deserted or taken objects or parked motor vehicles. It is really challenging for the existing approach to efficiently identify the objects as they occur. In this research suggested an effective method, Fuzzy based Ant Colony Optimization (FACO) integrated with modified kernel fuzzy c-means algorithm (MKFCM) for object segmentation. Here FACO algorithm situates optimal initial cluster centroid for the MKFCM, thus improve all applications affiliated fuzzy clustering such as foreground segmentation in image processing. The recommended new method is intellectual and additionally dynamic clustering technique for the dividing of motion objects. After the segmentation, the object tracking can be done using particle filtering method. The morphological operation helps the particle filter to effectively track the objects. Here the flow is based on the traditional waterfall approach. From the empirical impacts, the recommended method outshines than the state of artwork. Below the suggested method attains maximum efficiency for both PETS and also Hall monitor videos when examined to the current algorithm.

Keyword:

Modified kernel fuzzy c-means

Ant Colony Optimization (ACO)

Fuzzy logic (FL)

Foreground

Background

Clustering