To provide pest technicians with a convenient way to recognize insects,a novel method is proposed to classify insect images by integrated region matching (IRM) and dual tree complex wavelet transform (DTCWT).The wing ...To provide pest technicians with a convenient way to recognize insects,a novel method is proposed to classify insect images by integrated region matching (IRM) and dual tree complex wavelet transform (DTCWT).The wing image of the lepidopteran insect is preprocessed to obtain the region of interest (ROI) whose position is then calibrated.The ROI is first segmented with the k-means algorithm into regions according to the color features,properties of all the segmented regions being used as a coarse level feature.The color image is then converted to a grayscale image,where DTCWT features are extracted as a fine level feature.The IRM scheme is undertaken to find K nearest neighbors (KNNs),out of which the nearest neighbor is searched by computing the Canberra distance of DTCWT features.The method was tested with a database including 100 lepidopteran insect species from 18 families and the recognition accuracy was 84.47%.For the forewing subset,a recognition accuracy of 92.38% was achieved.The results showed that the proposed method can effectively solve the problem of automatic species identification of lepidopteran specimens.展开更多
Using the "matching extent" method,this article conducts a multi-perspective analysis on the compatibility of the investment strategies of the western region and MNCs,as well as national policies on foreign ...Using the "matching extent" method,this article conducts a multi-perspective analysis on the compatibility of the investment strategies of the western region and MNCs,as well as national policies on foreign investment. Research indicates that low matching extent amongst the above three decision-makers is an important reason for the disadvantage of the western region in terms of FDI attraction,and this situation will not change in the short term. In addition,this article also sheds light on how to increase matching extent.展开更多
基于PatchMatch的多视图立体(MVS)方法依据输入多幅图像估计场景的深度,目前已应用于大规模场景三维重建。然而,由于特征匹配不稳定、仅依赖光度一致性不可靠等原因,现有方法在弱纹理区域的深度估计准确性和完整性较低。针对上述问题,...基于PatchMatch的多视图立体(MVS)方法依据输入多幅图像估计场景的深度,目前已应用于大规模场景三维重建。然而,由于特征匹配不稳定、仅依赖光度一致性不可靠等原因,现有方法在弱纹理区域的深度估计准确性和完整性较低。针对上述问题,提出一种基于四叉树先验辅助的MVS方法。首先,利用图像像素值获得局部纹理;其次,基于自适应棋盘网格采样的块匹配多视图立体视觉方法(ACMH)获得粗略的深度图,结合弱纹理区域中的结构信息,采用四叉树分割生成先验平面假设;再次,融合上述信息,设计一种新的多视图匹配代价函数,引导弱纹理区域得到最优深度假设,进而提高立体匹配的准确性;最后,在ETH3D、Tanks and Temples和中国科学院古建筑数据集上与多种现有的传统MVS方法进行对比实验。结果表明所提方法性能更优,特别是在ETH3D测试数据集中,当误差阈值为2 cm时,相较于当前先进的多尺度平面先验辅助方法(ACMMP),它的F1分数和完整性分别提高了1.29和2.38个百分点。展开更多
针对行人被障碍物部分遮挡导致的检测准确率降低问题,提出了基于多特征融合的树形路径半全局立体匹配的部分遮挡行人检测算法。使用简单线性迭代聚类(simple linear iterative clustering,SLIC)算法进行超像素分割,提升行人的轮廓信息,...针对行人被障碍物部分遮挡导致的检测准确率降低问题,提出了基于多特征融合的树形路径半全局立体匹配的部分遮挡行人检测算法。使用简单线性迭代聚类(simple linear iterative clustering,SLIC)算法进行超像素分割,提升行人的轮廓信息,并使用多特征融合的树形路径半全局立体匹配算法生成深度图;对行人信息和背景信息及障碍物信息使用自适应分割算法进行分离,获取感兴趣区域;将感兴趣区域放置在行人特征明显且稳定的头肩部,进行感兴趣区域的约束;使用降维梯度直方图特征(histogram of gradient,HOG)进行特征提取并生成样本集,训练支持向量机(support vector machines,SVM)分类器,最终实现部分遮挡的行人检测。实验表明,所提算法与其他行人检测算法相比,在行人部分遮挡场景下,有着更高的行人检测准确率,证明所提算法的有效性。展开更多
In this paper,we propose a novel improved region energy based image fusion rule.The original images are firstly decomposed by using the lifting scheme of wavelet transform into four sub-bands:LL,LH,HL,HH,by studying p...In this paper,we propose a novel improved region energy based image fusion rule.The original images are firstly decomposed by using the lifting scheme of wavelet transform into four sub-bands:LL,LH,HL,HH,by studying principles and characteristics of the wavelet subbands,and we put emphasis on the high frequency subbands.Thus HH,HL,LH sub-bands,which represent three direction of high frequency details,are weighted by different size of three direction Gaussian kernel,then the energy based image fusion rule is applied with a optional size of window,thus the activity level of high frequency subbands are obtained,followed by a local region matching degree in the corresponding direction and resolution,an activity level of low frequency subband is calculated,then perform consistency verification on the selected wavelet coefficients,by doing the inverse wavelet transform the fused image is obtained.The performance of the proposed novel image fusion scheme is conducted and compared with a few existing image fusion algorithm,the experimental results show that the proposed method is an effective multi-focus image fusion algorithm.展开更多
针对图像特征匹配过程中由于杂乱的室外场景或待匹配目标被物体遮挡而产生的外点导致匹配精度低及鲁棒性差等问题,提出了一种融合基于密度的带噪空间聚类算法(Density Based Spatial Clustering of Applications with Noise,DBSCAN)与Tw...针对图像特征匹配过程中由于杂乱的室外场景或待匹配目标被物体遮挡而产生的外点导致匹配精度低及鲁棒性差等问题,提出了一种融合基于密度的带噪空间聚类算法(Density Based Spatial Clustering of Applications with Noise,DBSCAN)与Two-stage策略的改进LoFTR的图像特征匹配方法D2S-LoFTR。首先将原始图像1/8维度的特征匹配作为初始匹配结果,使用基于空间密度的DBSCAN算法对其特征进行聚类,提取最优匹配对的同时滤除由外点造成的误匹配。接着裁剪出由聚类得到的两幅原始图像中的共视区域,使用卷积层注意力模块(Convolutional Block Attention Module,CBAM)对其特征重构后进行二次匹配,将匹配结果与初始匹配进行融合以增强匹配的准确性。在室外数据集Megadepth上的实验结果表明,D2S-LoFTR的平均特征匹配率达到93.47%,与LoFTR相比提升1.91%,在旋转误差阈值为5°,10°,20°情况下的相对位姿估计累计曲线下面积(Area Under the cumulative Curve,AUC)分别为55.12%、71.03%、82.02%,分别提升2.32%,1.84%,0.84%,证实所提方法能够更好地适应杂乱室外场景下的图像特征匹配任务。展开更多
Height extraction for buildings is a fundamental step of 3D scene reconstruction in many virtual reality applications. In this paper, we propose an automatic method to extract the height of buildings in high resolutio...Height extraction for buildings is a fundamental step of 3D scene reconstruction in many virtual reality applications. In this paper, we propose an automatic method to extract the height of buildings in high resolution satellite imagery based on the length of shadow. Taking into account the limitation of traditional algorithms, we make use of the boundary information of a building to facilitate detecting and matching the shadow regions with higher accuracy. Then, we introduce a shadow-cast model to correct the shadow location in our system. The experimental result shows that when extracting the height of buildings from complex urban regions, our method has better accuracy.展开更多
基金Project (No.2006AA10Z211) supported by the National High-Tech Research and Development Program (863) of China
文摘To provide pest technicians with a convenient way to recognize insects,a novel method is proposed to classify insect images by integrated region matching (IRM) and dual tree complex wavelet transform (DTCWT).The wing image of the lepidopteran insect is preprocessed to obtain the region of interest (ROI) whose position is then calibrated.The ROI is first segmented with the k-means algorithm into regions according to the color features,properties of all the segmented regions being used as a coarse level feature.The color image is then converted to a grayscale image,where DTCWT features are extracted as a fine level feature.The IRM scheme is undertaken to find K nearest neighbors (KNNs),out of which the nearest neighbor is searched by computing the Canberra distance of DTCWT features.The method was tested with a database including 100 lepidopteran insect species from 18 families and the recognition accuracy was 84.47%.For the forewing subset,a recognition accuracy of 92.38% was achieved.The results showed that the proposed method can effectively solve the problem of automatic species identification of lepidopteran specimens.
文摘Using the "matching extent" method,this article conducts a multi-perspective analysis on the compatibility of the investment strategies of the western region and MNCs,as well as national policies on foreign investment. Research indicates that low matching extent amongst the above three decision-makers is an important reason for the disadvantage of the western region in terms of FDI attraction,and this situation will not change in the short term. In addition,this article also sheds light on how to increase matching extent.
文摘基于PatchMatch的多视图立体(MVS)方法依据输入多幅图像估计场景的深度,目前已应用于大规模场景三维重建。然而,由于特征匹配不稳定、仅依赖光度一致性不可靠等原因,现有方法在弱纹理区域的深度估计准确性和完整性较低。针对上述问题,提出一种基于四叉树先验辅助的MVS方法。首先,利用图像像素值获得局部纹理;其次,基于自适应棋盘网格采样的块匹配多视图立体视觉方法(ACMH)获得粗略的深度图,结合弱纹理区域中的结构信息,采用四叉树分割生成先验平面假设;再次,融合上述信息,设计一种新的多视图匹配代价函数,引导弱纹理区域得到最优深度假设,进而提高立体匹配的准确性;最后,在ETH3D、Tanks and Temples和中国科学院古建筑数据集上与多种现有的传统MVS方法进行对比实验。结果表明所提方法性能更优,特别是在ETH3D测试数据集中,当误差阈值为2 cm时,相较于当前先进的多尺度平面先验辅助方法(ACMMP),它的F1分数和完整性分别提高了1.29和2.38个百分点。
文摘针对行人被障碍物部分遮挡导致的检测准确率降低问题,提出了基于多特征融合的树形路径半全局立体匹配的部分遮挡行人检测算法。使用简单线性迭代聚类(simple linear iterative clustering,SLIC)算法进行超像素分割,提升行人的轮廓信息,并使用多特征融合的树形路径半全局立体匹配算法生成深度图;对行人信息和背景信息及障碍物信息使用自适应分割算法进行分离,获取感兴趣区域;将感兴趣区域放置在行人特征明显且稳定的头肩部,进行感兴趣区域的约束;使用降维梯度直方图特征(histogram of gradient,HOG)进行特征提取并生成样本集,训练支持向量机(support vector machines,SVM)分类器,最终实现部分遮挡的行人检测。实验表明,所提算法与其他行人检测算法相比,在行人部分遮挡场景下,有着更高的行人检测准确率,证明所提算法的有效性。
基金Sponsored by the National Natural Science Foundation of China(Grant No.61077079)the Ph.D.Programs Foundation of Ministry of Education of China(Grant No.20102304110013)+1 种基金the Key Program of Heilongjiang Natural Science Foundation(Grant No.ZD201216)the Program ExcellentAcademic Leaders of Harbin(Grant No.RC2013XK009003)
文摘In this paper,we propose a novel improved region energy based image fusion rule.The original images are firstly decomposed by using the lifting scheme of wavelet transform into four sub-bands:LL,LH,HL,HH,by studying principles and characteristics of the wavelet subbands,and we put emphasis on the high frequency subbands.Thus HH,HL,LH sub-bands,which represent three direction of high frequency details,are weighted by different size of three direction Gaussian kernel,then the energy based image fusion rule is applied with a optional size of window,thus the activity level of high frequency subbands are obtained,followed by a local region matching degree in the corresponding direction and resolution,an activity level of low frequency subband is calculated,then perform consistency verification on the selected wavelet coefficients,by doing the inverse wavelet transform the fused image is obtained.The performance of the proposed novel image fusion scheme is conducted and compared with a few existing image fusion algorithm,the experimental results show that the proposed method is an effective multi-focus image fusion algorithm.
文摘针对图像特征匹配过程中由于杂乱的室外场景或待匹配目标被物体遮挡而产生的外点导致匹配精度低及鲁棒性差等问题,提出了一种融合基于密度的带噪空间聚类算法(Density Based Spatial Clustering of Applications with Noise,DBSCAN)与Two-stage策略的改进LoFTR的图像特征匹配方法D2S-LoFTR。首先将原始图像1/8维度的特征匹配作为初始匹配结果,使用基于空间密度的DBSCAN算法对其特征进行聚类,提取最优匹配对的同时滤除由外点造成的误匹配。接着裁剪出由聚类得到的两幅原始图像中的共视区域,使用卷积层注意力模块(Convolutional Block Attention Module,CBAM)对其特征重构后进行二次匹配,将匹配结果与初始匹配进行融合以增强匹配的准确性。在室外数据集Megadepth上的实验结果表明,D2S-LoFTR的平均特征匹配率达到93.47%,与LoFTR相比提升1.91%,在旋转误差阈值为5°,10°,20°情况下的相对位姿估计累计曲线下面积(Area Under the cumulative Curve,AUC)分别为55.12%、71.03%、82.02%,分别提升2.32%,1.84%,0.84%,证实所提方法能够更好地适应杂乱室外场景下的图像特征匹配任务。
基金Supported by National Natural Science Foundation of China(61232014,61421062,61472010)the National Key Technology R&D Program of China(2015BAK01B06)
文摘Height extraction for buildings is a fundamental step of 3D scene reconstruction in many virtual reality applications. In this paper, we propose an automatic method to extract the height of buildings in high resolution satellite imagery based on the length of shadow. Taking into account the limitation of traditional algorithms, we make use of the boundary information of a building to facilitate detecting and matching the shadow regions with higher accuracy. Then, we introduce a shadow-cast model to correct the shadow location in our system. The experimental result shows that when extracting the height of buildings from complex urban regions, our method has better accuracy.
