Determination of body posture invariant relationships for identifying and tracking human | |
| Author | Duta, Arunav |
| Call Number | AIT Thesis no.ISE-08-15 |
| Subject(s) | Image processing Posture |
| Note | A Thesis submitted in partial fulfillment of the re quirements for the degree of Master of Engineering in Mechatronics, School of Engineering and Technology |
| Publisher | Asian Institute of Technology |
| Series Statement | Thesis ; no. ISE-08-15 |
| Abstract | Human detection has a variety of applications. There has been a large amount of work on it. The large variability shown by the humans, especially due to the various possible postures, is a major challenge to the process of detection. This thesis proposes a human detection system that is expected to exhibit invariance with regard to the posture of the human. The system uses three tests to detect a human in an image, which are the Hausdorff test, Moment test and Skin test. The hausdorff and the moment test require the creation of two databases, namely the human database and the non-human database. The three tests are carried out on the test images consecutively. The hausdorff test determines the Hausdorff distance between the test image and the images of the databases and then classified. The Moment test compares the test image and the database images based on Hu's seven invariant moment set. The Hausdorff and the Moment tests use 3-NN (Nearest Neighbor) classifier. The Skin filter converts the images to the normalized RGB space and then looks for skin pixels in the images according to a specified criterion. A human is said to be detected by the system if any two filters used in the thesis produce positive results. Experimental results shown in the thesis demonstrate the accuracy and robustness of the system. |
| Year | 2008 |
| Corresponding Series Added Entry | Asian Institute of Technology. Thesis ; no. ISE-08-15 |
| Type | Thesis |
| School | School of Engineering and Technology (SET) |
| Department | Department of Industrial Systems Engineering (DISE) |
| Academic Program/FoS | Industrial Systems Engineering (ISE) |
| Chairperson(s) | Afzulpurkar, Nitin V. |
| Examination Committee(s) | Manukid Parnichkun;Thavida Maneewarn |
| Scholarship Donor(s) | Asian Institute of Technology Fellowship |
| Degree | Thesis (M.Eng.) - Asian Institute of Technology, 2008 |