In China, volatile organic compound（VOC） control directives have been continuously released and implemented for important sources and regions to tackle air pollution. The corresponding control requirements were base...In China, volatile organic compound（VOC） control directives have been continuously released and implemented for important sources and regions to tackle air pollution. The corresponding control requirements were based on VOC emission amounts（EA）, but never considered the significant differentiation of VOC species in terms of atmospheric chemical reactivity. This will adversely influence the effect of VOC reduction on air quality improvement. Therefore,this study attempted to develop a comprehensive classification method for typical VOC sources in the Beijing–Tianjin–Hebei region（BTH）, by combining the VOC emission amounts with the chemical reactivities of VOC species. Firstly, we obtained the VOC chemical profiles by measuring 5 key sources in the BTH region and referencing another 10 key sources, and estimated the ozone formation potential（OFP） per ton VOC emission for these sources by using the maximum incremental reactivity（MIR） index as the characteristic of source reactivity（SR）. Then, we applied the data normalization method to respectively convert EA and SR to normalized EA（NEA） and normalized SR（NSR） for various sources in the BTH region.Finally, the control index（CI） was calculated, and these sources were further classified into four grades based on the normalized CI（NCI）. The study results showed that in the BTH region,furniture coating, automobile coating, and road vehicles are characterized by high NCI and need to be given more attention; however, the petro-chemical industry, which was designated as an important control source by air quality managers, has a lower NCI.展开更多
China is a mountainous country with a great diversity of landform and geomorphology.This diversity underlines the need for regionalization and classification.This study defines the mountain terrains and regions with t...China is a mountainous country with a great diversity of landform and geomorphology.This diversity underlines the need for regionalization and classification.This study defines the mountain terrains and regions with three criteria-elevation,relative height,and slope,and examines the extent of mountainous regions by using county as the basic administrative unit.According to the three parameters of economic base,resident income and development potential,we classified the economic development level in mountainous regions of China.The findings reveal that the extent of the mountainous region accounts for 74.9% of the China's Mainland's total area.The economic development of mountainous regions in China is classified into 4 main types and 23 subtypes.展开更多
感兴趣区域(region of interest,ROI)图像的提取在运动目标的检测与跟踪等领域有着广泛的应用;HSV(hue saturation value)是根据颜色的直观特性创建的一种颜色空间,用于对颜色进行定量描述。文章将HSV颜色空间用于图像中颜色信息的计算...感兴趣区域(region of interest,ROI)图像的提取在运动目标的检测与跟踪等领域有着广泛的应用;HSV(hue saturation value)是根据颜色的直观特性创建的一种颜色空间,用于对颜色进行定量描述。文章将HSV颜色空间用于图像中颜色信息的计算,识别颜色特征。针对RoboMaster全国大学生机器人大赛中云台实时识别和捕捉敌方移动机器人装甲这一基本任务,提出一种基于色彩特征的目标识别算法。首先对图像进行阈值分割,计算二值图像轮廓的形状描述参数,寻找图像中所有的高亮条形物体;再用HSV颜色空间根据ROI颜色判断ROI中物体是否为目标灯柱。在此之前,需要统计符合要求的目标灯柱像素的HSV参数及RGB参数的大致范围,再结合HSV中典型颜色对应参数范围表,确定算法所采用的参数范围。最终识别目标的判定准则是,在初步筛选所得感兴趣区域中,HSV及RGB参数处于经分析所得参数范围内的像素点占整个区域的比例大于某一设定阈值。经测试,该方法对于像素为640×480的一组图片,平均处理速度可达到102.33帧/s,平均识别率约为97.4%。在RoboMaster大赛中,使用该目标识别算法的机器人能实时跟踪移动机器人,验证了该方法的有效性。展开更多
This paper proposes a solution to localization and classification of rice grains in an image.All existing related works rely on conventional based machine learning approaches.However,those techniques do not do well fo...This paper proposes a solution to localization and classification of rice grains in an image.All existing related works rely on conventional based machine learning approaches.However,those techniques do not do well for the problem designed in this paper,due to the high similarities between different types of rice grains.The deep learning based solution is developed in the proposed solution.It contains pre-processing steps of data annotation using the watershed algorithm,auto-alignment using the major axis orientation,and image enhancement using the contrast-limited adaptive histogram equalization(CLAHE)technique.Then,the mask region-based convolutional neural networks(R-CNN)is trained to localize and classify rice grains in an input image.The performance is enhanced by using the transfer learning and the dropout regularization for overfitting prevention.The proposed method is validated using many scenarios of experiments,reported in the forms of mean average precision(mAP)and a confusion matrix.It achieves above 80%mAP for main scenarios in the experiments.It is also shown to perform outstanding,when compared to human experts.展开更多
基金supported by the National Key Technology Support Program of China(Nos.2014BAC23B05&2014BAC23B02)the Natural Sciences Foundation of China(No.51478017)+1 种基金the Youth Individual Project of Beijing Talents Training Fund(No.2015000021733G170)the Ministry of Environmental Protection Special Funds for Scientific Research on Public Causes(No.201409016)
文摘In China, volatile organic compound（VOC） control directives have been continuously released and implemented for important sources and regions to tackle air pollution. The corresponding control requirements were based on VOC emission amounts（EA）, but never considered the significant differentiation of VOC species in terms of atmospheric chemical reactivity. This will adversely influence the effect of VOC reduction on air quality improvement. Therefore,this study attempted to develop a comprehensive classification method for typical VOC sources in the Beijing–Tianjin–Hebei region（BTH）, by combining the VOC emission amounts with the chemical reactivities of VOC species. Firstly, we obtained the VOC chemical profiles by measuring 5 key sources in the BTH region and referencing another 10 key sources, and estimated the ozone formation potential（OFP） per ton VOC emission for these sources by using the maximum incremental reactivity（MIR） index as the characteristic of source reactivity（SR）. Then, we applied the data normalization method to respectively convert EA and SR to normalized EA（NEA） and normalized SR（NSR） for various sources in the BTH region.Finally, the control index（CI） was calculated, and these sources were further classified into four grades based on the normalized CI（NCI）. The study results showed that in the BTH region,furniture coating, automobile coating, and road vehicles are characterized by high NCI and need to be given more attention; however, the petro-chemical industry, which was designated as an important control source by air quality managers, has a lower NCI.
基金supported by the National Natural Science Foundation of China(Grant No.41571523)the Key Program of the Chinese Academy of Sciences(Grant No.KZZDEW-06-03)+1 种基金the National Basic Research Program of China(Grant No.2013CBA01808)the National Key Technology R&D Program of China(Grant No.2014BAC05B01)
文摘China is a mountainous country with a great diversity of landform and geomorphology.This diversity underlines the need for regionalization and classification.This study defines the mountain terrains and regions with three criteria-elevation,relative height,and slope,and examines the extent of mountainous regions by using county as the basic administrative unit.According to the three parameters of economic base,resident income and development potential,we classified the economic development level in mountainous regions of China.The findings reveal that the extent of the mountainous region accounts for 74.9% of the China's Mainland's total area.The economic development of mountainous regions in China is classified into 4 main types and 23 subtypes.
文摘感兴趣区域(region of interest,ROI)图像的提取在运动目标的检测与跟踪等领域有着广泛的应用;HSV(hue saturation value)是根据颜色的直观特性创建的一种颜色空间,用于对颜色进行定量描述。文章将HSV颜色空间用于图像中颜色信息的计算,识别颜色特征。针对RoboMaster全国大学生机器人大赛中云台实时识别和捕捉敌方移动机器人装甲这一基本任务,提出一种基于色彩特征的目标识别算法。首先对图像进行阈值分割,计算二值图像轮廓的形状描述参数,寻找图像中所有的高亮条形物体;再用HSV颜色空间根据ROI颜色判断ROI中物体是否为目标灯柱。在此之前,需要统计符合要求的目标灯柱像素的HSV参数及RGB参数的大致范围,再结合HSV中典型颜色对应参数范围表,确定算法所采用的参数范围。最终识别目标的判定准则是,在初步筛选所得感兴趣区域中,HSV及RGB参数处于经分析所得参数范围内的像素点占整个区域的比例大于某一设定阈值。经测试,该方法对于像素为640×480的一组图片,平均处理速度可达到102.33帧/s,平均识别率约为97.4%。在RoboMaster大赛中,使用该目标识别算法的机器人能实时跟踪移动机器人,验证了该方法的有效性。
文摘This paper proposes a solution to localization and classification of rice grains in an image.All existing related works rely on conventional based machine learning approaches.However,those techniques do not do well for the problem designed in this paper,due to the high similarities between different types of rice grains.The deep learning based solution is developed in the proposed solution.It contains pre-processing steps of data annotation using the watershed algorithm,auto-alignment using the major axis orientation,and image enhancement using the contrast-limited adaptive histogram equalization(CLAHE)technique.Then,the mask region-based convolutional neural networks(R-CNN)is trained to localize and classify rice grains in an input image.The performance is enhanced by using the transfer learning and the dropout regularization for overfitting prevention.The proposed method is validated using many scenarios of experiments,reported in the forms of mean average precision(mAP)and a confusion matrix.It achieves above 80%mAP for main scenarios in the experiments.It is also shown to perform outstanding,when compared to human experts.
