为了进一步提高家居环境中异常行为的在线识别能力,将运动历史图像与表情识别相结合,提出了一种基于MHI(Motion History Image)与LBP(Local Binary Pattern)表情识别的辅助异常行为识别方法。该方法通过运动历史图像辅助识别运动轨迹,...为了进一步提高家居环境中异常行为的在线识别能力,将运动历史图像与表情识别相结合,提出了一种基于MHI(Motion History Image)与LBP(Local Binary Pattern)表情识别的辅助异常行为识别方法。该方法通过运动历史图像辅助识别运动轨迹,同时利用LBP提取人脸特征,进行面部表情识别,进而辅助识别异常行为状态。实验结果表明,本文方法对异常行为的识别率有一定的提升。展开更多
Backgrounds: Health does not only physical health, therefore, we need to study it from various viewpoints. Many Japanese female complain of a Hie or a low back pain (LBP), which they reduce their subjective well-being...Backgrounds: Health does not only physical health, therefore, we need to study it from various viewpoints. Many Japanese female complain of a Hie or a low back pain (LBP), which they reduce their subjective well-being. We analyze. Those patients often have acupuncture therapy. In this study, we analyze the characteristics of Hie and LBP, the satisfaction level of alternative therapy and we pursue a tip to improve subjective well-being. Methods: Of 1000 women, Hie (+)/Hie (−) or LBP (+)/LBP (−), we compared their body temperature (BT) (axilla) and body mass index (BMI). Furthermore, the Chi test identified ten factors of “body” and seven “mind” information. Results: In the result of BT (axilla) while LBP indicated a significant difference. Both Hie and LBP showed difference in the opposite direction. Hie did now show such clear differences in “body” information. However, interestingly, all seven questions in the “mind” information showed statistical difference. Discussion and conclusion: One reason why those patients have acupuncture therapy may acupuncture therapy traditionally has not separate “mind” and “body” and it has the concept of “mind-body unity”. To improve subjective well-being, first we need to focus on “Mind” as well as “mind-body unity”. Mental-health support is important for patients with Hie or LBP to reduce physiological stress.展开更多
针对传统LBP(Local Binary Pattern)算法在DR图像缺陷检测中对噪声异常敏感而导致的缺陷识别率低的问题,在已有的韦伯LBP算法(Weber Local Binary Pattern,WLBP)的基础上,提出改进的WALBP(Weber Adapted Local Binary Patterns)算法。WA...针对传统LBP(Local Binary Pattern)算法在DR图像缺陷检测中对噪声异常敏感而导致的缺陷识别率低的问题,在已有的韦伯LBP算法(Weber Local Binary Pattern,WLBP)的基础上,提出改进的WALBP(Weber Adapted Local Binary Patterns)算法。WALBP算法保留了WLBP算法最后生成二维直方图的特点,对其所用的LBP算子和Lo G(Laplacian of Gaussian)方法进行了改进。WALBP算法更加有效地描述了DR图像的纹理特征,同时有效解决了WLBP算子在进行缺陷检测时直方图维数较多及分类能力不强的问题。通过对多幅铸件DR图像进行实验分析,结果表明,相对于已有的WLBP算法和传统的LBP算法,WALBP算法在缺陷检测上具有更高的识别率,在缺陷识别技术中具有很高的应用价值。展开更多
文摘为了进一步提高家居环境中异常行为的在线识别能力,将运动历史图像与表情识别相结合,提出了一种基于MHI(Motion History Image)与LBP(Local Binary Pattern)表情识别的辅助异常行为识别方法。该方法通过运动历史图像辅助识别运动轨迹,同时利用LBP提取人脸特征,进行面部表情识别,进而辅助识别异常行为状态。实验结果表明,本文方法对异常行为的识别率有一定的提升。
文摘Backgrounds: Health does not only physical health, therefore, we need to study it from various viewpoints. Many Japanese female complain of a Hie or a low back pain (LBP), which they reduce their subjective well-being. We analyze. Those patients often have acupuncture therapy. In this study, we analyze the characteristics of Hie and LBP, the satisfaction level of alternative therapy and we pursue a tip to improve subjective well-being. Methods: Of 1000 women, Hie (+)/Hie (−) or LBP (+)/LBP (−), we compared their body temperature (BT) (axilla) and body mass index (BMI). Furthermore, the Chi test identified ten factors of “body” and seven “mind” information. Results: In the result of BT (axilla) while LBP indicated a significant difference. Both Hie and LBP showed difference in the opposite direction. Hie did now show such clear differences in “body” information. However, interestingly, all seven questions in the “mind” information showed statistical difference. Discussion and conclusion: One reason why those patients have acupuncture therapy may acupuncture therapy traditionally has not separate “mind” and “body” and it has the concept of “mind-body unity”. To improve subjective well-being, first we need to focus on “Mind” as well as “mind-body unity”. Mental-health support is important for patients with Hie or LBP to reduce physiological stress.
文摘针对传统LBP(Local Binary Pattern)算法在DR图像缺陷检测中对噪声异常敏感而导致的缺陷识别率低的问题,在已有的韦伯LBP算法(Weber Local Binary Pattern,WLBP)的基础上,提出改进的WALBP(Weber Adapted Local Binary Patterns)算法。WALBP算法保留了WLBP算法最后生成二维直方图的特点,对其所用的LBP算子和Lo G(Laplacian of Gaussian)方法进行了改进。WALBP算法更加有效地描述了DR图像的纹理特征,同时有效解决了WLBP算子在进行缺陷检测时直方图维数较多及分类能力不强的问题。通过对多幅铸件DR图像进行实验分析,结果表明,相对于已有的WLBP算法和传统的LBP算法,WALBP算法在缺陷检测上具有更高的识别率,在缺陷识别技术中具有很高的应用价值。
