法向量作为点云不可或缺的属性之一,在诸多算法中具有重要作用,因此法向估计一直是处理点云的一个重要任务。然而,对于尖锐特征点,其邻域是两个表面或者多个表面的交汇处,这会增加法向估计的难度。在处理尖锐特征点时,现有的一些算法仅...法向量作为点云不可或缺的属性之一,在诸多算法中具有重要作用,因此法向估计一直是处理点云的一个重要任务。然而,对于尖锐特征点,其邻域是两个表面或者多个表面的交汇处,这会增加法向估计的难度。在处理尖锐特征点时,现有的一些算法仅仅考虑了点和平面之间的距离属性,在法向估计时仍存在一些缺陷。为了更好地改善上述问题,Wang等人通过加权平面拟合的方法,不仅考虑点和平面之间的距离属性,而且加入了法向属性,该方法提高了法向估计的准确性。然而该方法并未考虑点和平面之间的法向属性。因此,本文在该方法的框架之上,考虑了点和拟合平面的法向差。实验结果表明,本文算法在法向估计的准确性上相比现有的算法有所提高。As one of the indispensable attributes of point cloud, normal vector plays an important role in many algorithms, so normal estimation has always been an important task in processing point cloud. However, for sharp feature points, their neighborhood is the intersection of two or more surfaces, which increases the difficulty of normal estimation. When processing sharp feature points, some existing algorithms only consider the distance attribute between the point and the plane, and there are still some defects in normal estimation. In order to better improve the above problems, Wang et al. not only consider the distance attribute between the point and the plane, but also add the normal attribute by the weighted plane fitting method, which improves the accuracy of normal estimation. However, this method does not consider the normal attribute between the point and the plane. Therefore, this paper considers the normal difference between the point and the fitting plane on the framework of this method. Experimental results show that the accuracy of normal estimation of the proposed algorithm is improved compared with the existing algorithms.展开更多
目的探讨骨形态发生蛋白-4(bone morphogenetic protein 4,BMP-4)基因对人乳腺癌MCF-7细胞增殖、凋亡和迁移能力的影响途径。方法通过向乳腺癌MCF-7细胞株转染PET-BMP-4质粒过表达BMP-4;采用RT-PCR、MTT法、Boyden小室培养、Hoechst染色...目的探讨骨形态发生蛋白-4(bone morphogenetic protein 4,BMP-4)基因对人乳腺癌MCF-7细胞增殖、凋亡和迁移能力的影响途径。方法通过向乳腺癌MCF-7细胞株转染PET-BMP-4质粒过表达BMP-4;采用RT-PCR、MTT法、Boyden小室培养、Hoechst染色、Western blot方法,检测质粒转染、NF-κB抑制剂吡咯二硫代氨基甲酸盐(pyrrodlidine dthiocarbamate,PDTC)预处理和对照组细胞增殖、凋亡、迁移能力及NF-κB、Bax、Bcl-2水平。结果与空质粒转染组和空白对照组细胞相比,过表达BMP-4基因可导致MCF-7细胞增殖能力升高,凋亡细胞比例降低,细胞迁移能力增强,NF-κB和Bcl-2表达水平升高,Bax表达水平降低(P<0.05);PDTC预处理后NF-κB和Bcl-2表达水平降低,Bax表达水平升高,显著降低过表达BMP-4对MCF-7细胞的生物学效应(P<0.05)。结论过表达BMP-4基因可增强乳腺癌MCF-7细胞的增殖和迁移能力并抑制其凋亡,可能与上调NF-κB有关。展开更多
在点云处理过程中,法向估计是非常重要的一步。现有的深度霍夫变换的法向估计网络通过对点云进行霍夫变换得到邻域特征,再将其输入至卷积神经网络中学习估计法向。但由于霍夫变换过程中存在一定信息损失导致最后所得法向不准确,效果不...在点云处理过程中,法向估计是非常重要的一步。现有的深度霍夫变换的法向估计网络通过对点云进行霍夫变换得到邻域特征,再将其输入至卷积神经网络中学习估计法向。但由于霍夫变换过程中存在一定信息损失导致最后所得法向不准确,效果不够理想。对此,本文先通过霍夫变换将法向空间与二维平面相对应,并将二维空间离散化获得所有潜在切平面,设计特征聚合将点特征转化为潜在切平面特征作为CNN输入来降低霍夫变换过程中信息的损失,从而提升卷积神经网络的输入,进而提升网络整体的法向估计质量。实验结果表明,由此产生的法向估计网络的整体性能有所提升,对于不同噪声尺度也更具鲁棒性。In the process of point cloud processing, normal estimation is a very important step. The existing normal estimation network of deep Hough transform obtains neighborhood features by performing Hough transform on the point cloud and then inputs them into a convolutional neural network to learn and estimate the normal. However, due to certain information loss in the Hough transform process, the finally obtained normal is inaccurate and the effect is not ideal. In response to this, this paper first corresponds the normal space to a two-dimensional plane through Hough transform, and discretizes the two-dimensional space to obtain all potential tangent planes. Feature aggregation is designed to transform point features into potential tangent plane features as the input of CNN to reduce the information loss in the Hough transform process, thereby enhancing the input of the convolutional neural network and further improving the overall normal estimation quality of the network. Experimental results show that the overall performance of the resulting normal estimation network is improved, and it is also more robust to different noise scales.展开更多
文摘法向量作为点云不可或缺的属性之一,在诸多算法中具有重要作用,因此法向估计一直是处理点云的一个重要任务。然而,对于尖锐特征点,其邻域是两个表面或者多个表面的交汇处,这会增加法向估计的难度。在处理尖锐特征点时,现有的一些算法仅仅考虑了点和平面之间的距离属性,在法向估计时仍存在一些缺陷。为了更好地改善上述问题,Wang等人通过加权平面拟合的方法,不仅考虑点和平面之间的距离属性,而且加入了法向属性,该方法提高了法向估计的准确性。然而该方法并未考虑点和平面之间的法向属性。因此,本文在该方法的框架之上,考虑了点和拟合平面的法向差。实验结果表明,本文算法在法向估计的准确性上相比现有的算法有所提高。As one of the indispensable attributes of point cloud, normal vector plays an important role in many algorithms, so normal estimation has always been an important task in processing point cloud. However, for sharp feature points, their neighborhood is the intersection of two or more surfaces, which increases the difficulty of normal estimation. When processing sharp feature points, some existing algorithms only consider the distance attribute between the point and the plane, and there are still some defects in normal estimation. In order to better improve the above problems, Wang et al. not only consider the distance attribute between the point and the plane, but also add the normal attribute by the weighted plane fitting method, which improves the accuracy of normal estimation. However, this method does not consider the normal attribute between the point and the plane. Therefore, this paper considers the normal difference between the point and the fitting plane on the framework of this method. Experimental results show that the accuracy of normal estimation of the proposed algorithm is improved compared with the existing algorithms.
文摘在点云处理过程中,法向估计是非常重要的一步。现有的深度霍夫变换的法向估计网络通过对点云进行霍夫变换得到邻域特征,再将其输入至卷积神经网络中学习估计法向。但由于霍夫变换过程中存在一定信息损失导致最后所得法向不准确,效果不够理想。对此,本文先通过霍夫变换将法向空间与二维平面相对应,并将二维空间离散化获得所有潜在切平面,设计特征聚合将点特征转化为潜在切平面特征作为CNN输入来降低霍夫变换过程中信息的损失,从而提升卷积神经网络的输入,进而提升网络整体的法向估计质量。实验结果表明,由此产生的法向估计网络的整体性能有所提升,对于不同噪声尺度也更具鲁棒性。In the process of point cloud processing, normal estimation is a very important step. The existing normal estimation network of deep Hough transform obtains neighborhood features by performing Hough transform on the point cloud and then inputs them into a convolutional neural network to learn and estimate the normal. However, due to certain information loss in the Hough transform process, the finally obtained normal is inaccurate and the effect is not ideal. In response to this, this paper first corresponds the normal space to a two-dimensional plane through Hough transform, and discretizes the two-dimensional space to obtain all potential tangent planes. Feature aggregation is designed to transform point features into potential tangent plane features as the input of CNN to reduce the information loss in the Hough transform process, thereby enhancing the input of the convolutional neural network and further improving the overall normal estimation quality of the network. Experimental results show that the overall performance of the resulting normal estimation network is improved, and it is also more robust to different noise scales.
