[Objective] We aimed to improve the quality of flue-cured tobacco with K326 upper leaves and explore the optimal tobacco baking process in Chongqing tobacco-producing region. [Method] Four different baking processes ...[Objective] We aimed to improve the quality of flue-cured tobacco with K326 upper leaves and explore the optimal tobacco baking process in Chongqing tobacco-producing region. [Method] Four different baking processes (endogenetic force moisture-removing flue-curing technique, exogenetic force moisture-removing flue-curing technique, three-stage flue-curing process, three-stage and six-step flue- curing process) were adopted in this research to investigate the effects of different baking processes on carotenoid content of same-maturity degree upper leaves of the flue-cured tobacco K326. [Result] The results showed at the yellowing and color fixing stages, the carotenoids were most fully decomposed in endogenetic force moisture-removing flue-curing technique. The decomposed carotenoid content was lowest in exogenetic force moisture-removing flue-curing technique, for which the carotenoid content was highest (50.53 mg/g). [Conclusion] At the yellowing and color fixing stages, exogenetic force moisture-removing flue-curing technique was most conducive to the carotenoid accumulation in tobacco baking process.展开更多
利用图像处理和模式识别技术进行复杂背景下黄瓜叶部病害的自动识别,需要先把目标叶片从复杂背景中分割出来,才能进行后续的特征提取和病害识别。为实现复杂背景下黄瓜叶片的分割,首先利用K-均值聚类算法去除图片中的非绿色部分,再采用...利用图像处理和模式识别技术进行复杂背景下黄瓜叶部病害的自动识别,需要先把目标叶片从复杂背景中分割出来,才能进行后续的特征提取和病害识别。为实现复杂背景下黄瓜叶片的分割,首先利用K-均值聚类算法去除图片中的非绿色部分,再采用基于laplacian of gaussia(LOG)算子的方法对待分割的叶片进行区域检测,然后进行基于形状上下文(shape context)的模板匹配和分割。为了提高匹配速度,先检测叶片的生长点和叶尖,以确定叶片的位置、尺寸和方向;然后使用基于超像素(superpixel)的最优匹配搜索方法来减少搜索的复杂度。对20幅黄瓜叶部病害图像进行分割测试,并与人工分割法进行对比,结果表明,本文所采用的分割算法能较好地从复杂背景下提取出黄瓜叶部病害图像,分割准确率达94.7%,为后期黄瓜病斑的特征提取等工作奠定了良好的基础。展开更多
基金Supported by General Program of Science and Technology Research of China National Tobacco Corporation([2012]No.122)Science and Technology Research Program of Chongqing Branch of China National Tobacco Corporation(NY20110601070010)Technology Project of Chinese Tobacco Corporation~~
文摘[Objective] We aimed to improve the quality of flue-cured tobacco with K326 upper leaves and explore the optimal tobacco baking process in Chongqing tobacco-producing region. [Method] Four different baking processes (endogenetic force moisture-removing flue-curing technique, exogenetic force moisture-removing flue-curing technique, three-stage flue-curing process, three-stage and six-step flue- curing process) were adopted in this research to investigate the effects of different baking processes on carotenoid content of same-maturity degree upper leaves of the flue-cured tobacco K326. [Result] The results showed at the yellowing and color fixing stages, the carotenoids were most fully decomposed in endogenetic force moisture-removing flue-curing technique. The decomposed carotenoid content was lowest in exogenetic force moisture-removing flue-curing technique, for which the carotenoid content was highest (50.53 mg/g). [Conclusion] At the yellowing and color fixing stages, exogenetic force moisture-removing flue-curing technique was most conducive to the carotenoid accumulation in tobacco baking process.
文摘利用图像处理和模式识别技术进行复杂背景下黄瓜叶部病害的自动识别,需要先把目标叶片从复杂背景中分割出来,才能进行后续的特征提取和病害识别。为实现复杂背景下黄瓜叶片的分割,首先利用K-均值聚类算法去除图片中的非绿色部分,再采用基于laplacian of gaussia(LOG)算子的方法对待分割的叶片进行区域检测,然后进行基于形状上下文(shape context)的模板匹配和分割。为了提高匹配速度,先检测叶片的生长点和叶尖,以确定叶片的位置、尺寸和方向;然后使用基于超像素(superpixel)的最优匹配搜索方法来减少搜索的复杂度。对20幅黄瓜叶部病害图像进行分割测试,并与人工分割法进行对比,结果表明,本文所采用的分割算法能较好地从复杂背景下提取出黄瓜叶部病害图像,分割准确率达94.7%,为后期黄瓜病斑的特征提取等工作奠定了良好的基础。