70results about How to "Avoid matching errors" patented technology

The invention discloses a pedestrian re-identification method. The method comprises steps of pedestrian characteristic extraction and pedestrian characteristic tolerance, for pedestrian characteristic extraction, image color histograms are extracted through employing sliding windows, main colors are expanded, statistics of appearance frequency of each color mode is carried out at each sliding window, sum of relatively larger frequencies is taken as characteristic output of the color, traversal of an integral image is carried out through the sliding windows, and characteristics are formed after normalization; for more generalization matching, a search source is extracted from characteristics of a full body to a half body; for tolerance calculation, characteristic dimensions with a non-zero search image characteristic vector and to-be-searched-target characteristics are taken as an Euclidean distance, and interference of background factors under a complex scene is further reduced. The method is advantaged in that re-identification under the complex scene is carried out, properties of strong transportability for scene change, stable algorithm and fast speed are realized, a problem of poor quality of monitoring videos and to-be-searched databases can be effectively solved, and strong practicality is realized.

The invention is suitable for the technical field of offline map matching, and provides an offline map matching method and device based on turning points and terminal equipment, and the method comprises the steps: obtaining a moving track of a floating vehicle, i.e., N GPS positioning points based on timestamp sorting; identifying turning points in the moving track to divide the moving track intoM sub-trajectories; performing matching according to the GPS positioning points at the two ends of the M sub-tracks to obtain a starting node and a destination node of each sub-track; constructing a candidate matching sub-network based on the mth sub-track segment; utilizing a candidate matching sub-network corresponding to the mth sub-track to perform matching calculation on the mth sub-track toobtain K shortest paths from the starting node to the destination node as K candidate sub-paths matched with the mth sub-track; and selecting an optimal matching sub-path for each sub-track, and finally connecting the M optimal matching sub-paths to obtain a final matching path matched with the moving track. According to the invention, matching efficiency and matching precision can be improved.

The invention discloses a multi-platform point cloud matching method based on surface features, and belongs to the technical field of mobile measurement point cloud matching. The method comprises thefollowing implementation steps: acquiring point cloud data of a target object, performing data denoising preprocessing, and converting a preprocessed original point cloud coordinate system into a local coordinate system which takes a gravity center as an origin of coordinates and has an invariable coordinate axis direction; uniformly selecting three pairs or more than three pairs of matching planes from the acquired point cloud data, performing plane fitting by adopting a robust RANSAC algorithm to extract plane parameters (A, B, C and D) of a common plane, and performing flatness evaluation on the plane point cloud; based on the plane parameters of the three pairs or more than three pairs of corresponding planes, converting the rotation parameters into a non-linear model, then performinglinearization processing, keeping the translation parameters unchanged, and performing parameter solving by adopting an overall least square method; evaluating the correlation of the to-be-registeredplanes based on the correction number of the adjustment result, and reselecting the planes which do not meet the requirements; and completing point cloud matching based on the coordinate conversion model.