Pendeteksi kelelahan mata pengemudi kendaraan menggunakan metode segmentasi warna dalam ruang warna YCBCR

Abstract

The progress and technology in development of transportation is increasing. However, this is also accompanied by the emergence of some undesirable negative effects such as increased number of traffic accidents. The increase in number of accidents is usually caused by various factors including human factors, vehicle factors, and environmental factors. The human factor is one of the most frequent factors causing traffic accidents. This system is designed as the manufacture of detection software for detect the condition of eyestrain in the driver of the vehicle using a camera connected to the computer as an image input device and measured performance of system develeopment. The method used in this system is to detect faces with segmentation RGB color to YCBCR color, eye detect with Roberts edge and the last method of simple logic as a classification of eye conditions. The system shows the results of classification with the highest accuracy is on video 1 of 85.40% and the lowest accuracy in video 7 is 13.67% whereas, the highest accuracy warning results on video 5 with 94.4% and the least accurate accuracy of warning with 25.26%.

Keywords: YCBCR color space segmentation, eye fatigue, eye classification system