Centralized storage and identity identification methods pose many risks,including hacker attacks,data misuse,and single points of failure.Additionally,existing centralized identity management methods face interoperabi...Centralized storage and identity identification methods pose many risks,including hacker attacks,data misuse,and single points of failure.Additionally,existing centralized identity management methods face interoperability issues and rely on a single identity provider,leaving users without control over their identities.Therefore,this paper proposes a mechanism for identity identification and data sharing based on decentralized identifiers.The scheme utilizes blockchain technology to store the identifiers and data hashed on the chain to ensure permanent identity recognition and data integrity.Data is stored on InterPlanetary File System(IPFS)to avoid the risk of single points of failure and to enhance data persistence and availability.At the same time,compliance with World Wide Web Consortium(W3C)standards for decentralized identifiers and verifiable credentials increases the mechanism’s scalability and interoperability.展开更多
Many performance indices for parallel mechanism are put forward in the phase of dimensional synthesis,except for identifiability index,which determines the difficulty of kinematical calibration.If the dimensional para...Many performance indices for parallel mechanism are put forward in the phase of dimensional synthesis,except for identifiability index,which determines the difficulty of kinematical calibration.If the dimensional parameters are inappropriately selected,the existing methods for optimizing identifiability will not effectively work.Thus,with the aim of studying identifiability optimization in dimensional synthesis for 3-PRS mechanism,kinematics with structural errors is analyzed to provide theoretical bases for kinematical model.Then through a comparison of two 3-PRS mechanisms with different dimensional parameters,identifiability performance is proved to be necessary and feasible for optimization in the phase of dimensional design.Finally,an index δ is proposed to scale the identifiability performance.With the index,identifiability analysis and dimensional synthesis simulation in the whole workspace is completed.The index is verified to be correct and feasible,and based on the index,a procedure of dimensional synthesis,as well as an example set of non-dimensional parameters of 3-PRS mechanism,is proposed.The proposed identifiability index and design method can effectively introduce identifiability optimization into dimensional synthesis,and will obviously benefit later kinematical calibration.展开更多
In this paper an original method based on the link between a piecewise identifiability analysis and a piecewise numerical estimation is presented for estimating parameters of a phenomenological diesel engine combustio...In this paper an original method based on the link between a piecewise identifiability analysis and a piecewise numerical estimation is presented for estimating parameters of a phenomenological diesel engine combustion model. This model is used for design, validation and pre-tuning of engine control laws. A cascade algebro-differential elimination method is used for studying identifiability. This investigation is done by using input-output-parameter relationship. Then these relations are transformed by using iterated integration. They are combined with an original numerical derivative estimation based on distribution theory which gives explicit point-wise derivative?estimation formulas for each given order. Then new approximate relations, linking block of parameters and outputs (without derivative) are obtained. These relations are linear relatively to the blocks of parameters and yield a first estimation of parameters which is used as initial guess for a local optimization method (least square method and a local search genetic algorithm).展开更多
Nowadays,wood identification is made by experts using hand lenses,wood atlases,and field manuals which take a lot of cost and time for the training process.The quantity and species must be strictly set up,and accurate...Nowadays,wood identification is made by experts using hand lenses,wood atlases,and field manuals which take a lot of cost and time for the training process.The quantity and species must be strictly set up,and accurate identification of the wood species must be made during exploitation to monitor trade and enforce regulations to stop illegal logging.With the development of science,wood identification should be supported with technology to enhance the perception of fairness of trade.An automatic wood identification system and a dataset of 50 commercial wood species from Asia are established,namely,wood anatomical images collected and used to train for the proposed model.In the convolutional neural network(CNN),the last layers are usually soft-max functions with dense layers.These layers contain the most parameters that affect the speed model.To reduce the number of parameters in the last layers of the CNN model and enhance the accuracy,the structure of the model should be optimized and developed.Therefore,a hybrid of convolutional neural network and random forest model(CNN-RF model)is introduced to wood identification.The accuracy’s hybrid model is more than 98%,and the processing speed is 3 times higher than the CNN model.The highest accuracy is 1.00 in some species,and the lowest is 0.92.These results show the excellent adaptability of the hybrid model in wood identification based on anatomical images.It also facilitates further investigations of wood cells and has implications for wood science.展开更多
近年来,波段选择在高光谱图像降维处理中得到了广泛地应用,然而常用的数据降维方法并没能将与人类视觉系统相关的信息进行有效利用,如果将人类与生俱来的视觉注意机制能力应用到高光谱图像中目标的视觉显著性特征的增强或识别,对于高光...近年来,波段选择在高光谱图像降维处理中得到了广泛地应用,然而常用的数据降维方法并没能将与人类视觉系统相关的信息进行有效利用,如果将人类与生俱来的视觉注意机制能力应用到高光谱图像中目标的视觉显著性特征的增强或识别,对于高光谱图像的目标检测研究无疑会产生相当的促进作用。研究提出引入视觉注意机制理论应用于波段选择研究,构建面向目标检测应用的视觉注意机制波段选择模型。通过分析计算波段图幅的目标与背景的可识别程度,量化所在波段对地物目标与背景的判别能力,提出了基于目标视觉可识别度的波段选择方法;利用LC显著性算法进行空间域的视觉显著性目标分析,计算背景与目标的显著性差异绝对值,提出基于LC显著目标结构分布的波段选择方法。将这两种方法结合提出的改进子空间划分方法,建立面向目标检测的视觉注意机制波段选择模型,并经高光谱遥感AVIRIS San Diego公开数据集进行目标检测实验验证,结果表明所提出的基于视觉注意机制的波段选择模型对于目标检测应用具有较好的检测效果,实现了数据降维和高效的计算处理。展开更多
为解决光线遮蔽、藻萍干扰以及稻叶尖形状相似等复杂环境导致稻田杂草识别效果不理想问题,该研究提出一种基于组合深度学习的杂草识别方法。引入MSRCP(multi-scale retinex with color preservation)对图像进行增强,以提高图像亮度及对...为解决光线遮蔽、藻萍干扰以及稻叶尖形状相似等复杂环境导致稻田杂草识别效果不理想问题,该研究提出一种基于组合深度学习的杂草识别方法。引入MSRCP(multi-scale retinex with color preservation)对图像进行增强,以提高图像亮度及对比度;加入ViT分类网络去除干扰背景,以提高模型在复杂环境下对小目标杂草的识别性能。在YOLOv7模型中主干特征提取网络替换为GhostNet网络,并引入CA注意力机制,以增强主干特征提取网络对杂草特征提取能力及简化模型参数计算量。消融试验表明:改进后的YOLOv7模型平均精度均值为88.2%,较原YOLOv7模型提高了3.3个百分点,参数量减少10.43 M,计算量减少66.54×109次/s。识别前先经过MSRCP图像增强后,与原模型相比,改进YOLOv7模型的平均精度均值提高了2.6个百分点,光线遮蔽、藻萍干扰以及稻叶尖形状相似的复杂环境下平均精度均值分别提高5.3、3.6、3.1个百分点,加入ViT分类网络后,较原模型平均精度均值整体提升了4.4个百分点,光线遮蔽、藻萍干扰一级稻叶尖形状相似的复杂环境下的平均精度均值较原模型整体提升了6.2、6.1、5.7个百分点。ViT-改进YOLOv7模型的平均精度均值为92.6%,相比于YOLOv5s、YOLOXs、MobilenetV3-YOLOv7、YOLOv8和改进YOLOv7分别提高了11.6、10.1、5.0、4.2、4.4个百分点。研究结果可为稻田复杂环境的杂草精准识别提供支撑。展开更多
文摘Centralized storage and identity identification methods pose many risks,including hacker attacks,data misuse,and single points of failure.Additionally,existing centralized identity management methods face interoperability issues and rely on a single identity provider,leaving users without control over their identities.Therefore,this paper proposes a mechanism for identity identification and data sharing based on decentralized identifiers.The scheme utilizes blockchain technology to store the identifiers and data hashed on the chain to ensure permanent identity recognition and data integrity.Data is stored on InterPlanetary File System(IPFS)to avoid the risk of single points of failure and to enhance data persistence and availability.At the same time,compliance with World Wide Web Consortium(W3C)standards for decentralized identifiers and verifiable credentials increases the mechanism’s scalability and interoperability.
基金supported by National Natural Science Foundation of China (Grant No. 50775125)National Hi-tech Research and Development Program of China (863 Program,Grant No. 2007AA042003,No. 2007AA041901)
文摘Many performance indices for parallel mechanism are put forward in the phase of dimensional synthesis,except for identifiability index,which determines the difficulty of kinematical calibration.If the dimensional parameters are inappropriately selected,the existing methods for optimizing identifiability will not effectively work.Thus,with the aim of studying identifiability optimization in dimensional synthesis for 3-PRS mechanism,kinematics with structural errors is analyzed to provide theoretical bases for kinematical model.Then through a comparison of two 3-PRS mechanisms with different dimensional parameters,identifiability performance is proved to be necessary and feasible for optimization in the phase of dimensional design.Finally,an index δ is proposed to scale the identifiability performance.With the index,identifiability analysis and dimensional synthesis simulation in the whole workspace is completed.The index is verified to be correct and feasible,and based on the index,a procedure of dimensional synthesis,as well as an example set of non-dimensional parameters of 3-PRS mechanism,is proposed.The proposed identifiability index and design method can effectively introduce identifiability optimization into dimensional synthesis,and will obviously benefit later kinematical calibration.
文摘In this paper an original method based on the link between a piecewise identifiability analysis and a piecewise numerical estimation is presented for estimating parameters of a phenomenological diesel engine combustion model. This model is used for design, validation and pre-tuning of engine control laws. A cascade algebro-differential elimination method is used for studying identifiability. This investigation is done by using input-output-parameter relationship. Then these relations are transformed by using iterated integration. They are combined with an original numerical derivative estimation based on distribution theory which gives explicit point-wise derivative?estimation formulas for each given order. Then new approximate relations, linking block of parameters and outputs (without derivative) are obtained. These relations are linear relatively to the blocks of parameters and yield a first estimation of parameters which is used as initial guess for a local optimization method (least square method and a local search genetic algorithm).
文摘Nowadays,wood identification is made by experts using hand lenses,wood atlases,and field manuals which take a lot of cost and time for the training process.The quantity and species must be strictly set up,and accurate identification of the wood species must be made during exploitation to monitor trade and enforce regulations to stop illegal logging.With the development of science,wood identification should be supported with technology to enhance the perception of fairness of trade.An automatic wood identification system and a dataset of 50 commercial wood species from Asia are established,namely,wood anatomical images collected and used to train for the proposed model.In the convolutional neural network(CNN),the last layers are usually soft-max functions with dense layers.These layers contain the most parameters that affect the speed model.To reduce the number of parameters in the last layers of the CNN model and enhance the accuracy,the structure of the model should be optimized and developed.Therefore,a hybrid of convolutional neural network and random forest model(CNN-RF model)is introduced to wood identification.The accuracy’s hybrid model is more than 98%,and the processing speed is 3 times higher than the CNN model.The highest accuracy is 1.00 in some species,and the lowest is 0.92.These results show the excellent adaptability of the hybrid model in wood identification based on anatomical images.It also facilitates further investigations of wood cells and has implications for wood science.
文摘近年来,波段选择在高光谱图像降维处理中得到了广泛地应用,然而常用的数据降维方法并没能将与人类视觉系统相关的信息进行有效利用,如果将人类与生俱来的视觉注意机制能力应用到高光谱图像中目标的视觉显著性特征的增强或识别,对于高光谱图像的目标检测研究无疑会产生相当的促进作用。研究提出引入视觉注意机制理论应用于波段选择研究,构建面向目标检测应用的视觉注意机制波段选择模型。通过分析计算波段图幅的目标与背景的可识别程度,量化所在波段对地物目标与背景的判别能力,提出了基于目标视觉可识别度的波段选择方法;利用LC显著性算法进行空间域的视觉显著性目标分析,计算背景与目标的显著性差异绝对值,提出基于LC显著目标结构分布的波段选择方法。将这两种方法结合提出的改进子空间划分方法,建立面向目标检测的视觉注意机制波段选择模型,并经高光谱遥感AVIRIS San Diego公开数据集进行目标检测实验验证,结果表明所提出的基于视觉注意机制的波段选择模型对于目标检测应用具有较好的检测效果,实现了数据降维和高效的计算处理。
文摘为解决光线遮蔽、藻萍干扰以及稻叶尖形状相似等复杂环境导致稻田杂草识别效果不理想问题,该研究提出一种基于组合深度学习的杂草识别方法。引入MSRCP(multi-scale retinex with color preservation)对图像进行增强,以提高图像亮度及对比度;加入ViT分类网络去除干扰背景,以提高模型在复杂环境下对小目标杂草的识别性能。在YOLOv7模型中主干特征提取网络替换为GhostNet网络,并引入CA注意力机制,以增强主干特征提取网络对杂草特征提取能力及简化模型参数计算量。消融试验表明:改进后的YOLOv7模型平均精度均值为88.2%,较原YOLOv7模型提高了3.3个百分点,参数量减少10.43 M,计算量减少66.54×109次/s。识别前先经过MSRCP图像增强后,与原模型相比,改进YOLOv7模型的平均精度均值提高了2.6个百分点,光线遮蔽、藻萍干扰以及稻叶尖形状相似的复杂环境下平均精度均值分别提高5.3、3.6、3.1个百分点,加入ViT分类网络后,较原模型平均精度均值整体提升了4.4个百分点,光线遮蔽、藻萍干扰一级稻叶尖形状相似的复杂环境下的平均精度均值较原模型整体提升了6.2、6.1、5.7个百分点。ViT-改进YOLOv7模型的平均精度均值为92.6%,相比于YOLOv5s、YOLOXs、MobilenetV3-YOLOv7、YOLOv8和改进YOLOv7分别提高了11.6、10.1、5.0、4.2、4.4个百分点。研究结果可为稻田复杂环境的杂草精准识别提供支撑。
