Table 1 Efficiency of existing methods for object segmentation and object tracking compared to the proposed method
From: Tracking of moving human in different overlapping cameras using Kalman filter optimized
References | Important aspects | Knowledge and presumptions | Form model | Utilized technique | Particular task |
|---|---|---|---|---|---|
[2] | Combining edge and region data | Interactive initialization with constant velocity | Elliptical cones are joined together | The 3d model’s 2d projection | Human tracking |
[31] | Employs B-splines | Learn constant velocity and shape in space | recognized shape space | In-state sampling | Track shapes of objects in space |
[32] | State space in high dimensions | Search by state | Initial model is defined by the scaled prismatic model | testing several hypotheses | Flexible objects |
Sreekala et al. [29] | Chains of patches | Define the first patches | Straight limbs | Estimate the next patch’s likely location | Human limb tracking |
[36] | Greyscale pictures | Learn the history and the projection templates | People in silhouettes | Tracking and analyzing shapes | Monitoring of inhabitants |
Barawkar and Kumar [40] | One observer, permanent camera | Understanding the static context | Simulated human form | Following blobs | Human tracking |
[41] | identification of occlusions | Draw the first contour | – | Snake without scales | occlusion zones for tracks |
[39] | b-splines, with constant velocity | Automatically master the shape-space | eigenshapes | Kalman filter | Track human motion |
[14] | Color histogram | Location of the original area | – | Mean-shift histogram matching | Identify moving items |
This Work | tracking moving objects with human identity | Search by state | People moving using their appearance | Fuzzy Logic, Kalman filter | Tracking of Moving human |