摘要：【中文】本发明提供一种基于变异型粒子群算法的图像边缘检测方法及装置，能够提高边缘检测效果和检测效率。所述方法包括：S101，初始化粒子群、最大的迭代次数及图像边缘的灰度阈值；S102，根据图像中像素灰度值与灰度阈值之间的关系，计算粒子适应度，选择适应度大的粒子进入下一代；S103，根据当前迭代次数动态调整变异概率，根据调整后的变异概率，对选择的粒子执行变异操作，并更新粒子的位置；S104，判断当前迭代次数是否达到最大的迭代次数，若是，则获取粒子群最优解得出对应的灰度阈值，否则，返回执行S102；S105，根据得到的粒子群最优解对应的灰度阈值进行图像边缘检测。本发明涉及图像处理技术领域。 【EN】The invention provides an image edge detection method and device based on a variant particle swarm algorithm, which can improve the edge detection effect and detection efficiency. The method comprises the following steps: s101, initializing a particle swarm, the maximum iteration times and a gray level threshold of an image edge; s102, calculating the particle fitness according to the relation between the pixel gray value and the gray threshold value in the image, and selecting the particles with high fitness to enter the next generation; s103, dynamically adjusting the mutation probability according to the current iteration times, executing mutation operation on the selected particles according to the adjusted mutation probability, and updating the positions of the particles; s104, judging whether the current iteration times reach the maximum iteration times, if so, obtaining the optimal solution of the particle swarm to obtain a corresponding gray threshold, and if not, returning to execute S102; and S105, detecting the image edge according to the gray threshold corresponding to the obtained particle swarm optimal solution. The invention relates to the technical field of image processing.

摘要：【中文】本发明提供一种基于中医医案的个性化知识图谱的构建方法及装置，能够直接反映出医生的行医风格和临床就诊思路。所述方法包括：获取目标中医的医案，以医案号作为关联，构建目标中医的医案知识图谱G；对医案知识图谱G按照类别间节点对进行遍历，提取节点对，以其中一个节点作为路径的出发节点S，另一个节点作为路径的终止节点E，在预先构建的中医基础理论知识图库提取出发节点为S、终止节点为E的路径p，组成路径集合C；将路径集合C中的所有路径，添加到医案知识图谱G中，得到所述目标中医的个性化医案知识图谱。本发明涉及中医学技术领域。 【EN】The invention provides a method and a device for constructing a personalized knowledge map based on a traditional Chinese medical scheme, which can directly reflect the doctor's travel style and clinical treatment thought. The method comprises the following steps: acquiring a medical record of a target traditional Chinese medicine, and constructing a medical record knowledge map G of the target traditional Chinese medicine by taking a medical record number as an association; traversing the medical case knowledge graph G according to node pairs among classes, extracting node pairs, taking one node as a starting node S of a path and the other node as a terminating node E of the path, extracting a path p with the starting node S and the terminating node E from a pre-constructed basic knowledge graph library of traditional Chinese medicine to form a path set C; and adding all the paths in the path set C into the medical case knowledge graph G to obtain the personalized medical case knowledge graph of the target traditional Chinese medicine. The invention relates to the technical field of traditional Chinese medicine.

摘要：【中文】本发明提供一种基于面向对象影像分类技术的蓝色屋顶建筑物提取方法，能够将蓝色屋顶建筑物从复杂的遥感影像中完整准确地提取出来。所述方法包括：获取对研究区域的遥感影像进行分割后得到的地物对象，其中，所述地物对象包括：蓝色屋顶建筑物；根据蓝色屋顶建筑物遥感影像的光谱特征和形状特征，得到用于将蓝色屋顶建筑物区分出的提取规则；利用提取规则，从分割结果中提取研究区域内的蓝色屋顶建筑物。本发明适用于遥感影像分类技术领域。 【EN】The invention provides a blue roof building extraction method based on an object-oriented image classification technology, which can completely and accurately extract a blue roof building from a complex remote sensing image. The method comprises the following steps: obtaining a surface feature object obtained by segmenting a remote sensing image of a research area, wherein the surface feature object comprises: a blue roof building; obtaining an extraction rule for distinguishing the blue roof buildings according to the spectral characteristics and the shape characteristics of the remote sensing images of the blue roof buildings; and extracting the blue roof buildings in the research area from the segmentation result by using an extraction rule. The method is suitable for the technical field of remote sensing image classification.

摘要：【中文】本发明提供一种基于U‑net改进算法的遥感影像变化检测方法及系统，该方法包括：对遥感影像进行预处理，然后将预处理后的不同时相的遥感影像进行整合得到输入图像；基于改进的U‑net网络，对输入图像进行下采样编码和上采样解码；最后根据分析结果输出影像是否发生变化的二值图像。本发明将语义分割方向的网络结构运用于变化检测领域，并引入残差学习机制使得编码器可快速收敛且加深网络层数，同时使用ASPP加强网络对图像特征的感知能力，使得算法在精度和效率上都能保持在较高水平，同时具有较强的鲁棒性。适用于遥感影像的变化检测领域的同时也可推广到其他领域，具有重要意义。 【EN】The invention provides a remote sensing image change detection method and a system based on a U-net improved algorithm, wherein the method comprises the following steps: preprocessing the remote sensing images, and then integrating the preprocessed remote sensing images in different time phases to obtain an input image; performing down-sampling coding and up-sampling decoding on an input image based on the improved U-net network; and finally, outputting a binary image of whether the image changes according to the analysis result. The invention applies the network structure of the semantic segmentation direction to the field of change detection, introduces a residual learning mechanism to enable an encoder to quickly converge and deepen the network layer number, and simultaneously uses the ASPP to enhance the perception capability of the network to the image characteristics, so that the algorithm can be kept at a higher level in both precision and efficiency and has stronger robustness. The method is suitable for the change detection field of remote sensing images, can be popularized to other fields, and has important significance.