Object Detection and Tracking on Three Dimensional Images Based-on a New Multishape Search- Pattern
Magdi B. M. Amien1, Alia M. A. Sidig2, Raza K. Yusif3
1Magdi B. M. Amien, Dept. of Electronics Engineering & Technology, University of Gezira, Sudan.
2Alia M. A. Sidig, Dept. of Electronics Engineering & Technology, University of Gezira, Sudan.
3Razaz K. Yusif, Dept. of Electronics Engineering & Technology, University of Gezira, Sudan.
Manuscript received on November 25, 2013. | Revised Manuscript received on December 15, 2013. | Manuscript published on December 30, 2013. | PP: 450-453 | Volume-3, Issue-2, December 2013. | Retrieval Number: B2515123213/2013©BEIESP
Open Access | Ethics and Policies | Cite
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Object detection and motion estimation are important issues in many different fields. They are widely and comprehensively used in military, robot industry, movie technology, medical field, and others. Therefore they have been the motivation of many research activities, through image and video processing. Among tens of available literature a number of approaches have been tried, but Block Matching, Optical Flow, and Block Flow are the famous techniques. This study introduces a new framework to deal with object detection and trajectory tracking problem, in a sequences of 3D ultrasound frames; firstly the traditional Block-Matching algorithm has been modified into a new multishape-search-pattern, and then we use combination of the modified-model and optical flow algorithm in a “cascade” to detect and determine the trajectory of the interested object. Atrial septal defect (ASD) has been selected as an object of case-study, and 3D ultrasound videos from “Khalifa-Hospital in Abu-Dhabi” were used as a data set, to evaluate the performance of the implemented algorithm. Comparative results show that the proposed scheme has a significant improvement in detecting and tracking ASDs, in terms of Peak Signal to Noise Ratio (PSNR) and computing velocity.
Keywords: Block Matching, Computer vision technology, Objects-Detection.