5 results about "Standard map" patented technology

This invention discloses a cross-view positioning method based on semantic coarse localization and BEV fine registration, aiming to solve the problem of large-scale, high-precision global positioning under conditions of no initial position prior. The invention employs a two-stage processing architecture: First, hierarchical semantic information is extracted from multi-view images from the vehicle using a visual language model. Combined with heading constraints and a backtracking retrieval strategy, candidate regions are quickly screened in a standard map to achieve semantic coarse localization. Then, within the candidate regions, the vehicle-mounted BEV perception results are finely geometrically registered with the map template, and a dynamic weight fusion mechanism based on road and building responses is introduced. The fusion weights are adaptively adjusted according to the degree of support from building structures to complete the fine registration. This invention organically combines high-level semantic information with low-level geometric information, achieving high-precision and robust cross-view geolocation even in the absence of coarse position prior. It effectively solves the problems of easy confusion in pure geometric matching, low accuracy in pure semantic localization, and poor adaptability of fixed fusion methods, making it suitable for global positioning of unmanned systems in complex urban environments.

The invention provides a fusion map generation method and an intelligent car control method. The fused map generation method comprises the following steps: mapping original map data of a plurality of navigation modes into a plurality of logic sub-layers of a fused map, and generating standard map data in different logic sub-layers; when the navigation mode is the laser radar navigation mode, generating standard map data in the background sub-layer; and when the navigation mode is the linear navigation mode or the matrix two-dimensional code navigation mode, generating standard map data in the path sub-layer. Original map data from different sources are fused into unified standard map data in a plurality of logic sub-layers of a fused map through a mapping relation, isomerism is digested through the fused map, and the problem of complexity of an original system is solved. According to the method, the existing facts are respected, the data consistency and uniformity are realized, the universality and usability of the fused map are enhanced, and the adaptability of the fused map is improved.

The invention relates to the technical field of artificial intelligence and natural language processing, in particular to a standard text generation system based on artificial intelligence. The system comprises a standard map construction and analysis unit, a multi-source heterogeneous data cleaning and mapping unit, a hierarchical constraint generation unit, an instant reverse verification unit and a feedback correction iteration unit. A constraint tree is constructed according to the system analysis standard, and original materials are mounted through semantic matching; the core of the method is that hierarchical generation is executed by using a large language model, and a compliance residual signal is calculated by comparing original constraints; triggering directional rewriting according to the residual error, and optimizing a metadata constraint based on a verification failure event; according to the method, a long text is decomposed into node-level microscopic tasks, so that logic chaos of a large model due to context limitation is eliminated, and logic leakproofness of a generation process is ensured.

The application provides a map registration method and device and a computer readable storage medium, relates to the technical field of maps, and can improve the efficiency of map auditing. The method comprises the following steps: acquiring map data of a to-be-audited map and map data of a plurality of standard maps; the zoom levels of each standard map are different; for each standard map in the plurality of standard maps, a standard map registration model between the standard map and a corresponding adjacent standard map of the standard map is determined; the zoom level of the adjacent standard map is adjacent to the zoom level of the standard map, the standard map registration model is used to determine the coordinate value of the homonym of a target point, and the to-be-audited map is registered with a target standard map; the target standard map is any one of the plurality of standard maps.

The invention relates to the technical field of geographic information processing, and discloses a common map-oriented quality inspection method and system and a storage medium, and the method comprises the steps: obtaining a map to be subjected to quality inspection and N standard maps, and extracting geometric data, attribute data and topological data; performing data preprocessing on the map to be subjected to quality inspection and N standard maps; based on the N standard maps, screening a reference map; constructing geometric data indexes, attribute data indexes and topological data indexes of the map to be subjected to quality inspection and the standard map relative to the reference map; generating a geometric data index standard value, an attribute data index standard value and a topological data index standard value; respectively comparing the geometric data index, the attribute data index and the topological data index of the map to be subjected to quality inspection relative to the reference map with corresponding standard values, and generating a quality inspection result; according to the invention, the quality inspection efficiency is improved, and the objectivity and reliability of the quality inspection result are obviously improved.