4results about How to "Effective registration" patented technology

The invention relates to an intelligent geographic surveying and mapping data processing analysis method and system, and relates to the field of data processing, and the method comprises the steps: sequentially carrying out the preprocessing, data registration and data fusion of multi-source geographic surveying and mapping data, and generating a regional standard surveying and mapping image; performing salient feature extraction on the regional standard surveying and mapping image to obtain a local salient surveying and mapping image set; based on a convolutional neural network, performing deviation traversal comparison on the local significant surveying and mapping image set and a historical local significant surveying and mapping image set, and outputting a surveying and mapping deviation data set; and inputting the surveying and mapping deviation data set into a flood spreading prediction plug-in, and outputting a flood spreading prediction result. According to the method, the technical problems that the accuracy and efficiency of data processing are insufficient and the accuracy of flood spreading prediction is poor due to the fact that data fusion and feature extraction cannot be effectively carried out in an existing method can be solved, efficient data fusion and registration can be realized, and the flood spreading prediction accuracy is improved. And the accuracy and reliability of flood spreading prediction are improved.

The application belongs to the field of computer vision three-dimensional point cloud registration, and particularly relates to a point cloud registration method based on registration error zero norm minimization. * ; using l2 norm to quantify the overall fitting error of the fitting error of each relative point pair in three spatial dimensions, and according to the matching point pair corresponding to the minimum K r ; using l2 norm to quantify the overall fitting error of the fitting error of each relative point pair in three spatial dimensions, and according to the matching point pair corresponding to the minimum K * ; using l2 norm to quantify the overall fitting error of the fitting error of each relative point pair in three spatial dimensions, and according to the matching point pair corresponding to the minimum K * ; using l2 norm to quantify the overall fitting error of the fitting error of each relative point pair in three spatial dimensions, and according to the matching point pair corresponding to the minimum K t ; using l2 norm to quantify the overall fitting error of the fitting error of each relative point pair in three spatial dimensions, and according to the matching point pair corresponding to the minimum K * ; using l2 norm to quantify the overall fitting error of the fitting error of each relative * point pair in three spatial dimensions, and according to the matching point pair corresponding to the minimum K The application can realize efficient point cloud registration in the presence of a large number of outliers.

The invention discloses a BGR + LBP two-parameter description method for registration feature points in aerial images of a small unmanned aerial vehicle. The method comprises the following steps: acquiring BGR values of the feature points; constructing a feature point adjacent pixel 3 * 3 window; calculating a BGR mean value of the feature points and eight pixel points around the feature points; obtaining an adjacent pixel reference value of the feature point # imgabs0 #; and obtaining a BGR + LBP two-parameter description value of the feature point # imgabs1 #. According to the method, the color features of the feature points can be effectively displayed, strong constraint on the feature points can be achieved through the LBP values of the surrounding pixels, noise interference possibly existing around can be effectively reduced, and technical support can be provided for achieving efficient registration of the feature points in the later period and greatly reducing the calculation amount of registration of the feature points.

The present invention relates to the field of computer graphics technology, and solves the technical problem that the existing point cloud registration methods are inefficient when facing sparse point cloud data, and in particular, relates to a point cloud registration method based on a cyclic twin attention mechanism, including: obtaining three-dimensional point cloud data of adjacent perspectives in a target scene, namely, source point cloud data and target point cloud data; constructing a point cloud registration model for predicting the rotation and translation required for registration between target point cloud data and source point cloud data; determining a loss function and optimization parameters for improving the point cloud registration model to accurately predict the rotation and translation required for registration; based on the rotation and translation of the target point cloud data predicted by the point cloud registration model, efficient registration of adjacent perspective point clouds is achieved. The point cloud registration method proposed in the present invention achieves efficient registration of adjacent target point cloud data when facing sparse point cloud data.