全部专利 发明专利 实用新型专利 外观设计专利
排序方式:
当前查询到5306条专利包含查询词 "Jiao Licheng" ,搜索用时0.203125秒!
发明专利:2421实用新型: 2501外观设计: 384
2421 条,当前第 1-10 条 下一页 最后一页 返回搜索页
申请号:201911073842.9 公开号:CN110874593A 主分类号:G06K9/32
申请人:【中文】西安电子科技大学【EN】XIDIAN University 申请日:2019.11.06 公开日:2020.03.10
摘要:【中文】本发明提出了一种基于掩膜的遥感图像旋转目标检测方法,旨在降低遥感图像旋转目标检测模型的计算量,并提高遥感图像旋转目标检测精度,实现步骤为:1)获取训练样本和测试样本;2)构建遥感图像旋转目标检测网络模型;3)用训练样本对遥感图像旋转目标检测网络模型进行训练;4)将测试样本输入已经训练好的遥感图像旋转目标检测网络模型中,获取遥感图像旋转目标的预测类别和旋转边界框四个顶点坐标。本发明通过目标旋转边界框确定遥感图像旋转目标的位置,有效减少了旋转目标检测模型计算量,增强了旋转目标定位鲁棒性,实现了更高的检测精度,可用于资源勘探、自然灾害预警、城市规划等领域。 【EN】The invention provides a mask-based remote sensing image rotating target detection method, which aims to reduce the calculated amount of a remote sensing image rotating target detection model and improve the detection precision of the remote sensing image rotating target and comprises the following steps: 1) acquiring a training sample and a test sample; 2) constructing a remote sensing image rotating target detection network model; 3) training a remote sensing image rotating target detection network model by using a training sample; 4) inputting the test sample into the trained remote sensing image rotating target detection network model, and obtaining the prediction category of the remote sensing image rotating target and the coordinates of four vertexes of the rotating boundary box. The method determines the position of the remote sensing image rotating target through the target rotating boundary frame, effectively reduces the calculated amount of a rotating target detection model, enhances the positioning robustness of the rotating target, realizes higher detection precision, and can be used in the fields of resource exploration, natural disaster early warning, urban planning and the like.
详细信息 下载全文

申请号:201911087534.1 公开号:CN110853057A 主分类号:G06T7/11
申请人:【中文】西安电子科技大学【EN】XIDIAN University 申请日:2019.11.08 公开日:2020.02.28
摘要:【中文】本发明公开一种基于全局和多尺度全卷积网络的航拍图像分割方法,其步骤为:构建全局和多尺度全卷积网络;生成训练集;训练全局和多尺度全卷积网络;将待分割的航拍图像输入到训练好的全局和多尺度全卷积网络进行二值分割,生成分割掩码图。本发明利用全局和多尺度全卷积网络对航拍图像进行分割,并在全局和多尺度全卷积网络中嵌入全局模块和多尺度模块,提取更加精细的分割掩码,鲁棒性强,分割精度高。 【EN】The invention discloses an aerial image segmentation method based on global and multi-scale full convolution networks, which comprises the following steps: constructing a global and multi-scale full convolution network; generating a training set; training a global and multi-scale full convolution network; and inputting the aerial image to be segmented into a trained global and multi-scale full convolution network for binary segmentation to generate a segmentation mask image. The method utilizes the global and multi-scale full convolution network to segment the aerial image, and embeds the global module and the multi-scale module in the global and multi-scale full convolution network, thereby extracting more refined segmentation mask, having strong robustness and high segmentation precision.
详细信息 下载全文

申请号:201911097973.0 公开号:CN110909161A 主分类号:G06F16/35
申请人:【中文】西安电子科技大学【EN】XIDIAN University 申请日:2019.11.12 公开日:2020.03.24
摘要:【中文】一种基于密度聚类和视觉相似度的英文单词分类方法,其步骤为:英文单词预处理;计算所选英文单词与每个英文单词的视觉相似度和距离;若邻域集中单词个数大于或等于2,则将所选单词加入空簇后选择未分类英文单词;处理未访问的未分类英文单词;对已访问的未分类英文单词,直接将其加入簇;判断是否有未分类英文单词,若有则选择未分类英文单词,否则将簇作为新一类单词标记为已访问;若单词均已访问,输出各类。本发明通过计算英文单词之间的视觉相似度与距离,并利用距离进行密度聚类,可以将视觉相似的英文单词分别组成一类,以提高对英文单词的记忆。 【EN】An English word classification method based on density clustering and visual similarity comprises the following steps: preprocessing English words; calculating the visual similarity and distance between the selected English word and each English word; if the number of the words in the neighborhood set is more than or equal to 2, adding the selected words into the empty cluster, and then selecting the unclassified English words; processing the unclassified English words which are not accessed; directly adding the accessed unclassified English words into a cluster; judging whether unclassified English words exist or not, if yes, selecting the unclassified English words, and otherwise, marking the clusters as new words as accessed; if the words are all accessed, the categories are output. According to the invention, by calculating the visual similarity and distance between English words and performing density clustering by using the distance, the English words with similar visual similarity can be respectively combined into one class, so that the memory of the English words is improved.
详细信息 下载全文

申请号:201911166777.4 公开号:CN110993691A 主分类号:H01L29/78
申请人:【中文】西安电子科技大学【EN】XIDIAN University 申请日:2019.11.25 公开日:2020.04.10
摘要:【中文】本发明提出了一种双沟道横向超结双扩散金属氧化物宽带隙半导体场效应管及其制作方法。该器件基于宽带隙材料,漂移区为超结结构;在基区的上部临近超结漂移区的一端向外侧依次形成第一源区、沟道衬底接触、第二源区以及沟槽;器件表面对应于第一源区相应的第一沟道的区域形成平面栅绝缘层以及平面栅电极;沟槽贯穿基区并延伸到下方的缓冲层,沟槽的底面和侧面形成沟槽栅绝缘层,并基于沟槽栅绝缘层内表面以多晶硅填平形成沟槽栅电极;在第一源区、沟道衬底接触和第二源区表面短接形成源电极。本发明的平面栅和沟槽栅结构形成双栅结构,实现电子电流的双沟道导通,有效降低了器件的导通电阻。 【EN】The invention provides a double-channel transverse super-junction double-diffusion metal oxide wide band gap semiconductor field effect transistor and a manufacturing method thereof. The device is based on a wide band gap material, and a drift region is of a super junction structure; a first source region, a channel substrate contact, a second source region and a groove are sequentially formed on the outer side of one end, close to the super junction drift region, of the upper portion of the base region; forming a planar gate insulating layer and a planar gate electrode on the surface of the device corresponding to the region of the first channel corresponding to the first source region; the groove penetrates through the base region and extends to the lower buffer layer, a groove gate insulating layer is formed on the bottom surface and the side surface of the groove, and a groove gate electrode is formed by filling and leveling polycrystalline silicon on the basis of the inner surface of the groove gate insulating layer; and forming a source electrode in the first source region, the channel substrate contact and the surface short circuit of the second source region. The plane gate and the groove gate structure form a double-gate structure, double-channel conduction of electron current is realized, and the on-resistance of the device is effectively reduced.
详细信息 下载全文

申请号:201911166780.6 公开号:CN111081777A 主分类号:H01L29/78
申请人:【中文】西安电子科技大学【EN】XIDIAN University 申请日:2019.11.25 公开日:2020.04.28
摘要:【中文】本发明提出了一种双沟道横向超结双扩散金属氧化物元素半导体场效应管及其制作方法。该器件的特点是:基于超结结构,衬底的材料为元素半导体材料;在基区的上部临近超结漂移区的一端向外侧依次形成第一源区、沟道衬底接触、第二源区以及沟槽;器件表面对应于第一源区相应的第一沟道的区域形成平面栅绝缘层以及平面栅电极;沟槽贯穿基区并延伸到下方的缓冲层,沟槽的底面和侧面形成沟槽栅绝缘层,并基于沟槽栅绝缘层内表面以多晶硅填平形成沟槽栅电极;在第一源区、沟道衬底接触和第二源区表面短接形成源电极。本发明的平面栅和沟槽栅结构形成双栅结构,实现电子电流的双沟道导通,有效降低了器件的导通电阻。 【EN】The invention provides a double-channel transverse super-junction double-diffusion metal oxide semiconductor field effect transistor and a manufacturing method thereof. The device is characterized in that: based on the super junction structure, the substrate is made of an element semiconductor material; a first source region, a channel substrate contact, a second source region and a groove are sequentially formed on the outer side of one end, close to the super junction drift region, of the upper portion of the base region; forming a planar gate insulating layer and a planar gate electrode on the surface of the device corresponding to the region of the first channel corresponding to the first source region; the groove penetrates through the base region and extends to the lower buffer layer, a groove gate insulating layer is formed on the bottom surface and the side surface of the groove, and a groove gate electrode is formed by filling and leveling polycrystalline silicon on the basis of the inner surface of the groove gate insulating layer; and forming a source electrode in the first source region, the channel substrate contact and the surface short circuit of the second source region. The plane gate and the groove gate structure form a double-gate structure, double-channel conduction of electron current is realized, and the on-resistance of the device is effectively reduced.
详细信息 下载全文

申请号:201911171794.7 公开号:CN110929080A 主分类号:G06F16/583
申请人:【中文】西安电子科技大学【EN】XIDIAN University 申请日:2019.11.26 公开日:2020.03.27
摘要:【中文】本发明公开了一种基于注意力和生成对抗网络的光学遥感图像检索方法,主要解决现有技术中光学遥感图像检索精度低的问题。本发明具体步骤如下:(1)构建深度卷积网络;(2)构建注意力网络;(3)构建生成对抗网络;(4)构建哈希学习网络;(5)训练网络;(6)获取每幅光学遥感图像的哈希编码向量;(7)检索光学遥感图像。本发明构建注意力网络,提取图像的可判别性特征,提高了图像特征的表达能力;构建生成对抗网络,提取图像哈希编码向量,减小了量化误差;最终提高了光学遥感图像的检索精度。 【EN】The invention discloses an optical remote sensing image retrieval method based on attention and generation countermeasure network, which mainly solves the problem of low retrieval precision of optical remote sensing images in the prior art. The method comprises the following specific steps: (1) constructing a deep convolutional network; (2) constructing an attention network; (3) constructing and generating a confrontation network; (4) constructing a Hash learning network; (5) training a network; (6) obtaining a Hash coding vector of each optical remote sensing image; (7) and retrieving the optical remote sensing image. According to the invention, an attention network is constructed, the discriminability characteristics of the image are extracted, and the expression capability of the image characteristics is improved; a countermeasure network is constructed and generated, and the image hash coding vector is extracted, so that the quantization error is reduced; and finally, the retrieval precision of the optical remote sensing image is improved.
详细信息 下载全文

申请号:201911217909.1 公开号:CN110991532A 主分类号:G06K9/62
申请人:【中文】西安电子科技大学【EN】XIDIAN University 申请日:2019.12.03 公开日:2020.04.10
摘要:【中文】本发明公开了一种基于关系视觉注意机制的场景图产生方法,主要解决现有技术中冗余的关系预测与可解释性较差的问题。其实现方案是:1)通过目标检测得到图像中的目标的类别与边界框,并进行全连接关系图建立;2)通过分析数据集,对关系图进行稀疏化,得到稀疏关系图表示;3)通过交替迭代学习关系注意力转移函数,分别从主语、宾语依靠并集特征转移到发生关系处,学习到准确的关系表征;4)对于学习到的关系表征进行分类,并组合成最终的场景图。本发明利用两目标发生关系的内在联系,建立关系注意力机制准确地关注于发生关系的区域,实现了场景图的准确产生,提高了网络的可解释化性,可用于图像描述与视觉问答任务。 【EN】The invention discloses a scene graph generation method based on a relational visual attention mechanism, which mainly solves the problem of poor redundant relational prediction and interpretability in the prior art. The implementation scheme is as follows: 1) obtaining the category and the bounding box of the target in the image through target detection, and establishing a full connection relation graph; 2) sparsifying the relational graph by analyzing the data set to obtain a sparse relational graph representation; 3) learning accurate relation representation by alternately and iteratively learning a relation attention transfer function and respectively transferring the subject and the object to a relation generation position depending on the union set characteristics; 4) and classifying the learned relational representation and combining into a final scene graph. The method utilizes the internal relation of the occurrence relation of the two targets to establish a relation attention mechanism to accurately focus on the area of the occurrence relation, realizes the accurate generation of a scene graph, improves the interpretability of the network, and can be used for image description and visual question and answer tasks.
详细信息 下载全文

申请号:201911300840.9 公开号:CN111079649A 主分类号:G06K9/00
申请人:【中文】西安电子科技大学【EN】XIDIAN University 申请日:2019.12.17 公开日:2020.04.28
摘要:【中文】本发明公开了一种基于轻量化语义分割网络的遥感图像地物分类方法,主要解决现有方法由于图像空间和通道特征信息利用不足且模型庞大,而导致的对遥感图像地物分类精度不高、训练速度较慢的问题。其方案为:在遥感图像地物分类数据集中获取训练样本和测试样本;构建引入可拓宽通道分解空洞卷积的轻量化遥感图像地物分类模型,设计关注地物边缘的整体损失函数;将训练样本输入到所构建的地物分类模型中训练,得到训练好的模型;将测试样本输入训练好的模型中,预测输出遥感图像中地物分类结果。本发明提升了特征的表达能力,减少了网络参数,提高了遥感图像地物分类的平均精度和训练速度,可用于获取一幅遥感图像的地物分布情况。 【EN】The invention discloses a remote sensing image ground feature classification method based on a lightweight semantic segmentation network, which mainly solves the problems of low precision and low training speed of remote sensing image ground feature classification caused by insufficient utilization of image space and channel characteristic information and huge model in the existing method. The scheme is as follows: acquiring a training sample and a test sample in a remote sensing image ground object classification data set; constructing a lightweight remote sensing image ground object classification model introducing the convolution of the widened channel decomposition cavity, and designing an integral loss function concerning the edge of the ground object; inputting the training sample into the constructed ground feature classification model for training to obtain a trained model; and inputting the test sample into the trained model, and predicting and outputting the ground feature classification result in the remote sensing image. The invention improves the expression capability of the characteristics, reduces the network parameters, improves the average precision and the training speed of the ground feature classification of the remote sensing image, and can be used for obtaining the ground feature distribution condition of one remote sensing image.
详细信息 下载全文

申请号:201911338663.3 公开号:CN111027511A 主分类号:G06K9/00
申请人:【中文】西安电子科技大学【EN】XIDIAN University 申请日:2019.12.23 公开日:2020.04.17
摘要:【中文】本发明公开了一种基于感兴趣区块提取的光学遥感图像舰船检测方法,主要解决现有技术检测精度低,虚警较多的问题。其实现方案为:构建光学遥感图像舰船检测数据集;对宽幅遥感图像下采样并去雾增强,利用上下文信息和图像全局特征进行水陆分割;用构建的数据集训练基于SCRDet的目标检测模型;根据水陆分割结果,用部分重叠的滑动窗口扫描原宽幅遥感图像提取感兴趣区块作为待检测区域,将待检测区域图像输入检测模型得到区域检测结果;将区域结果映射到原宽幅图像尺度上,作改进的非极大值抑制以优化初步检测结果;根据舰船的结构特征再次优化检测结果。本发明检测精度高,虚警率低,可用于获取大幅面遥感图像中感兴趣的舰船目标及其的位置。 【EN】The invention discloses an optical remote sensing image ship detection method based on region of interest block extraction, which mainly solves the problems of low detection precision and more false alarms in the prior art. The implementation scheme is as follows: constructing an optical remote sensing image ship detection data set; downsampling and defogging enhancing the wide remote sensing image, and carrying out land and water segmentation by using context information and image global characteristics; training a target detection model based on the SCRDEt by using the constructed data set; according to the land and water segmentation result, scanning the original wide remote sensing image by using a partially overlapped sliding window to extract an interested area block as a to-be-detected area, and inputting the to-be-detected area image into a detection model to obtain an area detection result; mapping the region result to the original wide image scale, and performing improved non-maximum suppression to optimize the primary detection result; and optimizing the detection result again according to the structural characteristics of the ship. The method has high detection precision and low false alarm rate, and can be used for acquiring the ship target of interest and the position thereof in the large-format remote sensing image.
详细信息 下载全文

申请号:201911342809.1 公开号:CN111080652A 主分类号:G06T7/10
申请人:【中文】西安电子科技大学【EN】XIDIAN University 申请日:2019.12.23 公开日:2020.04.28
摘要:【中文】本发明公开了一种基于多尺度轻量化空洞卷积的光学遥感图像分割方法,主要解决现有技术中网络所占的存储空间大、图像分割效果欠佳的问题。其实现方案为:获取光学遥感图像数据,并划分训练样本集和测试样本集;构建由特征提取下采样子网络、底层子网络、图像恢复上采样子网络级联的多尺度轻量化空洞卷积网络;用训练样本集对构建的多尺度轻量化空洞卷积网络进行训练;将测试样本集输入到训练好的多尺度轻量化空洞卷积网络中进行测试,得到光学遥感图像的分割结果。本发明减小了分割网络所占存储空间,提高了对光学遥感图像的分割精度,可用于土地规划管理,植被资源调查及环境监测。 【EN】The invention discloses an optical remote sensing image segmentation method based on multi-scale lightweight void convolution, which mainly solves the problems of large storage space occupied by a network and poor image segmentation effect in the prior art. The implementation scheme is as follows: acquiring optical remote sensing image data, and dividing a training sample set and a test sample set; constructing a multi-scale lightweight cavity convolution network formed by cascading a feature extraction down-sampling sub-network, a bottom sub-network and an image recovery up-sampling sub-network; training the constructed multi-scale lightweight void convolution network by using a training sample set; and inputting the test sample set into a trained multi-scale lightweight void convolution network for testing to obtain a segmentation result of the optical remote sensing image. The invention reduces the storage space occupied by the segmentation network, improves the segmentation precision of the optical remote sensing image, and can be used for land planning management, vegetation resource investigation and environment monitoring.
详细信息 下载全文

2421 条,当前第 1-10 条 下一页 最后一页 返回搜索页
©2022 Patent9.com All rights reserved.