期刊名称：TELKOMNIKA (Telecommunication Computing Electronics and Control)
印刷版ISSN：2302-9293
出版年度：2015
卷号：13
期号：3
页码：949-954
DOI：10.12928/telkomnika.v13i3.713
语种：English
出版社：Universitas Ahmad Dahlan
摘要：Face recognition is one of early detection in security system. Automation encourages implementation of face recognition in robot with low-power processor. Most of face recognition research focused on recognition accuration only and performed on high-speed computer. Face recognition that is implemented on low-cost processor, such as ARM processor, needs proper algorithm. Our research proposed K-Nearest Neighbor (KNN) algorithm to recognize face by ARM processor, which was common processor in robot system. This research sought best k-value to create proper face recognition with low-power processor. The proposed algorithm was tested on AT&T face dataset from Computer Laboratory, Cambridge University. The 15 images were set as testing image and 315 images were used as reference data set. OpenCV was choosen as main core image processing library, due to its high-speed. Proposed algorithm was implemented on ARM11 700MHz. Experiment result showed that KNN face recognition detected 93.3% face with k=1. Another proposed Histogram KNN face recognition gave 100% true detection with k=3. Overall experiment showed that proposed algorithm detected face on 2.657 s by ARM processor.
其他摘要：Face recognition is one of early detection in security system. Automation encourages implementation of face recognition in robot with low-power processor. Most of face recognition research focused on recognition accuration only and performed on high-speed computer. Face recognition that is implemented on low-cost processor, such as ARM processor, needs proper algorithm. Our research proposed K-Nearest Neighbor (KNN) algorithm to recognize face by ARM processor, which was common processor in robot system. This research sought best k-value to create proper face recognition with low-power processor. The proposed algorithm was tested on AT&T face dataset from Computer Laboratory, Cambridge University. The 15 images were set as testing image and 315 images were used as reference data set. OpenCV was choosen as main core image processing library, due to its high-speed. Proposed algorithm was implemented on ARM11 700MHz. Experiment result showed that KNN face recognition detected 93.3% face with k=1. Another proposed Histogram KNN face recognition gave 100% true detection with k=3. Overall experiment showed that proposed algorithm detected face on 2.657 s by ARM processor.
