Patents
Literature
Patsnap Eureka AI that helps you search prior art, draft patents, and assess FTO risks, powered by patent and scientific literature data.

2results about How to "Improve spatial precision" patented technology

A measurement method of a nanometer and micrometer non-contact in-situ optical thermal characterization technology combination

ActiveCN116448802BHigh precisionImproved longitudinal measurement accuracy
The application discloses a kind of nanometer non-contact in-situ optical thermal characterization technology combination measurement method, this method includes: for larger device size, measurement accuracy requirement is not high, qualitative evaluation main heat source area or thermal failure positioning wide band gap semiconductor power device thermal characterization application demand, using infrared thermal imaging mode;For smaller device size, measurement accuracy requirement is higher, quantitative characterization wide band gap, ultra-wide band gap semiconductor power device temperature distribution thermal characterization application demand, for transverse structure device using thermal reflection imaging mode and nanometer particle-based Raman thermal characterization mode characterization measurement device metal electrode temperature and device surface temperature;For vertical structure device using thermal reflection imaging mode, Raman thermal characterization mode and nanometer particle-based Raman thermal characterization mode characterization measurement device metal electrode temperature, device surface temperature and device surface temperature under even thickness.
Owner:XIDIAN UNIV

A Visual Language Navigation Method for Unmanned Aerial Vehicles Based on Predictive Semantic Occupation Representation

ActiveCN122083964AReduce invalid exploration milesshorten the time
This invention provides a UAV visual language navigation method based on predictive semantic occupancy representation, comprising: acquiring multimodal visual observations and six-degree-of-freedom pose data; extracting semantic features using a large visual language model and back-projecting them to construct a local three-dimensional semantic occupancy representation; performing target existence likelihood inference on the observation boundary and unexplored airspace based on the semantic prior of the large model to generate predictive semantic occupancy representations; parsing natural language commands to extract target semantic description vectors, aligning them with the predictive representations across modalities, and calculating semantic response scores; extracting candidate frontiers based on semantic response scores and obstacle avoidance constraints, calculating comprehensive exploration utility to select local navigation target points; performing local trajectory planning that satisfies kinematic constraints, driving the UAV to fly along a collision-free smooth trajectory, and updating the environmental representation online in a closed loop based on real physical observations until the mission is completed.
Owner:CHINA UNIV OF PETROLEUM (EAST CHINA)