目的为解决疲劳驾驶检测中人眼状态识别的难点,提出一种基于眼白分割的疲劳检测方法。方法首先对获取图像进行人脸检测,利用眼白在Cb-Cr上良好的聚类性,基于YCb Cr颜色空间建立高斯眼白分割模型;然后在人脸区域图像内做眼白分割,计算眼...目的为解决疲劳驾驶检测中人眼状态识别的难点,提出一种基于眼白分割的疲劳检测方法。方法首先对获取图像进行人脸检测,利用眼白在Cb-Cr上良好的聚类性,基于YCb Cr颜色空间建立高斯眼白分割模型;然后在人脸区域图像内做眼白分割,计算眼白面积;最后将眼白面积作为人眼开度指标,结合PERCLOS(percentage of eyelid closure over the pupil over time)判定人的疲劳状态。结果选取10个短视频进行采帧分析,实验结果表明,高斯眼白分割模型能有效分离眼白,并识别人眼开合状态,准确率可达96.77%。结论在良好光线条件下,本文方法能取得不错的分割效果;本文所提出的以眼白面积作为判定人眼开度的指标,能准确地判定人的疲劳状态。实验结果证明了该方法的有效性,值得今后做更深入的研究。展开更多
The influence of different driving cycles on their exhaust emissions and fuel consumption rate of gasoline passenger car was investigated in Bangkok based on the actual measurements obtained from a test vehicle drivin...The influence of different driving cycles on their exhaust emissions and fuel consumption rate of gasoline passenger car was investigated in Bangkok based on the actual measurements obtained from a test vehicle driving on a standard chassis dynamometer. A newly established Bangkok driving cycle （BDC） and the European driving cycle （EDC） which is presently adopted as the legislative cycle for testing automobiles registered in Thailand were used. The newly developed BDC is constructed using the driving characteristic data obtained from the real on-road driving tests along selected traffic routes. A method for selecting appropriate road routes for real driving tests is also introduced. Variations of keyed driving parameters of BDC with different driving cycles were discussed. The results showed that the HC and CO emission factors of BDC are almost two and four times greater than those of EDC, respectively. Although the difference in the NOx emission factor is small, the value from BDC is still greater than that of EDC by 10%. Under BDC, the test vehicle consumes fuel about 25% more than it does under EDC. All these differences are mainly attributed to the greater proportion of idle periods and higher fluctuations of vehicle speed in the BDC cycle. This result indicated that the exhausted emissions and fuel consumption of vehicles obtained from tests under the legislative modal-type driving cycle （EDC） are significantly different from those actually produced under real traffic conditions especially during peak periods.展开更多
文摘目的为解决疲劳驾驶检测中人眼状态识别的难点,提出一种基于眼白分割的疲劳检测方法。方法首先对获取图像进行人脸检测,利用眼白在Cb-Cr上良好的聚类性,基于YCb Cr颜色空间建立高斯眼白分割模型;然后在人脸区域图像内做眼白分割,计算眼白面积;最后将眼白面积作为人眼开度指标,结合PERCLOS(percentage of eyelid closure over the pupil over time)判定人的疲劳状态。结果选取10个短视频进行采帧分析,实验结果表明,高斯眼白分割模型能有效分离眼白,并识别人眼开合状态,准确率可达96.77%。结论在良好光线条件下,本文方法能取得不错的分割效果;本文所提出的以眼白面积作为判定人眼开度的指标,能准确地判定人的疲劳状态。实验结果证明了该方法的有效性,值得今后做更深入的研究。
基金funded by the Energy Policyand Planning Office (EPPO) of Thailand
文摘The influence of different driving cycles on their exhaust emissions and fuel consumption rate of gasoline passenger car was investigated in Bangkok based on the actual measurements obtained from a test vehicle driving on a standard chassis dynamometer. A newly established Bangkok driving cycle （BDC） and the European driving cycle （EDC） which is presently adopted as the legislative cycle for testing automobiles registered in Thailand were used. The newly developed BDC is constructed using the driving characteristic data obtained from the real on-road driving tests along selected traffic routes. A method for selecting appropriate road routes for real driving tests is also introduced. Variations of keyed driving parameters of BDC with different driving cycles were discussed. The results showed that the HC and CO emission factors of BDC are almost two and four times greater than those of EDC, respectively. Although the difference in the NOx emission factor is small, the value from BDC is still greater than that of EDC by 10%. Under BDC, the test vehicle consumes fuel about 25% more than it does under EDC. All these differences are mainly attributed to the greater proportion of idle periods and higher fluctuations of vehicle speed in the BDC cycle. This result indicated that the exhausted emissions and fuel consumption of vehicles obtained from tests under the legislative modal-type driving cycle （EDC） are significantly different from those actually produced under real traffic conditions especially during peak periods.
