历史建筑遥感监测识别数据
收藏资源简介：
用于实现对历史建筑遥感监测中历史建筑点位的高效精准识别，主要识别内容为历史建筑管理中的居住小区、工业遗产及公共建筑。算法通过历史建筑点位在影像中的相对位置，自动计算出现有历史建筑的实地坐标、面积等信息。为后续的人员管理与问题派发提供方便，有利于网格员及时掌握自己管理区域的情况。解决了网格员在传统人工巡查中难以发现问题，容易忽略问题，巡查范围太大，隐蔽地段、偏远地区与恶劣环境中不易巡查等问题。将系统识别出的问题点位派发给网格员，使网格员有依据、有目标的实地确认，极大的提高了发现问题的效率，节约人员时间与人工成本，避免网格员出现无效的巡查。基于无人机航拍采集的历史建筑遥感影像数据，通过YOLO算法进行实时目标检测。首先将单元神经网络应用于2024年1月的遥感影像，将图像分割成19x19的单元格，每个单元神经网络负责预测K个单元格。预测每个区域的概率，所有单元格上具有最大概率的类被选择并分配给特定的网格单元，生成预测点坐标（x，y），坐标系为CGCS2000，坐标为东经、北纬。 在预测类概率后，进行NMS运算，来消除不必要的锚点。算法识别下一个最高类别概率的边界框，并进行相同的运算过程，直到剩下所有不同的边界框。算法输出所需的要素，并显示各个类的边界框的细节。 抽取部分样本进行识别准确度验证，小于0.6视为识别错误，显示为FALSE；一般样本的识别准确度在0.8至1之间，大于等于0.6视为识别正确，显示为TRUE。通过判断结果正确或错误来纳入或排除数据，将识别正确的点位判定为历史建筑类别。最后将纳入的点位坐标、建筑类型等信息自动上传至历史建筑智能监管平台，获得历史建筑遥感监测识别数据。
This dataset is designed for efficient and accurate identification of historical building locations in remote sensing monitoring of historical buildings, with the main identification targets including residential communities, industrial heritage sites and public buildings under historical building management. Based on the relative positions of historical building locations in remote sensing images, the algorithm automatically calculates the field coordinates, area and other information of existing historical buildings. This facilitates subsequent personnel management and issue assignment, and helps grid workers timely grasp the status of their managed areas. It solves the problems faced by grid workers in traditional manual inspections, such as difficulty in detecting issues, tendency to overlook problems, overly large inspection scope, and challenges in inspecting hidden, remote areas and harsh environments. By assigning the issue locations identified by the system to grid workers, it enables them to conduct on-site verification with clear basis and targets, greatly improving the efficiency of problem detection, saving personnel time and labor costs, and avoiding ineffective inspections by grid workers. Real-time object detection is carried out using the YOLO algorithm based on historical building remote sensing image data collected via UAV aerial photography. Firstly, the grid neural network is applied to the remote sensing images from January 2024, which divides the image into 19×19 grid cells, and each grid cell is responsible for predicting K bounding boxes. The probability of each region is predicted, and the class with the highest probability among all grid cells is selected and assigned to the corresponding grid unit, generating the predicted point coordinates (x, y) in the CGCS2000 coordinate system, where the coordinates are east longitude and north latitude. After predicting the class probabilities, Non-Maximum Suppression (NMS) is performed to eliminate redundant bounding boxes. The algorithm identifies the bounding box with the next highest class probability and repeats the same process until all distinct bounding boxes remain. Finally, the required elements are output, and the details of the bounding boxes for each class are displayed. Some samples are selected for recognition accuracy verification. A score lower than 0.6 is considered a recognition error, marked as FALSE; for general samples, the recognition accuracy ranges from 0.8 to 1, and a score greater than or equal to 0.6 is considered a correct recognition, marked as TRUE. Data is included or excluded based on whether the recognition result is correct or incorrect, and the correctly identified locations are classified as historical buildings. Finally, information such as the coordinates and building types of the included locations is automatically uploaded to the intelligent supervision platform for historical buildings, thereby obtaining the historical building remote sensing monitoring and recognition dataset.
