Visual Localization Map Preloading via Pose Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing visual localization map loading technologies face delays due to the time-consuming process of loading sub-map files, which affects the instantaneity of visual localization, especially when updating local maps on the fly without prior loading of key frames and map points into memory.
Innovation Solution
A method that predicts and loads sub-map files based on the current pose by using a master map file to determine the necessary group numbers, allowing for preloading of required sub-map files, thereby eliminating wait times and ensuring real-time visual localization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If sub-map files are loaded only when needed during visual localization, then memory overhead is reduced, but localization instantaneity deteriorates due to loading wait time
Solution Approach 1:
The system performs preliminary action by predicting which sub-map files will be needed based on the current pose and desired trajectory, and loads these files into memory in advance before they are actually needed for localization. This eliminates the loading wait time during localization while avoiding loading unnecessary data that would waste memory.
Solution Approach 2:
The system dynamically adjusts the loading strategy by continuously updating the predicted set of group numbers based on the current pose and desired trajectory. As the vehicle moves and the trajectory is updated, the prediction adapts to load the correct subset of sub-map files, optimizing both memory usage and localization speed in real-time.
2Speed
If all sub-map files are loaded into memory, then localization speed is improved, but memory overhead increases significantly
Solution Approach 1:
The visual localization map is segmented into multiple sub-map files, each storing map data of a specific group obtained by grouping the visual localization map based on key frames. This segmentation allows the system to load only the necessary portions into memory rather than the entire map, reducing memory overhead while maintaining fast access to required data.
Solution Approach 2:
The system performs preliminary prediction to identify which segmented sub-map files will be needed based on the current pose and desired trajectory, and pre-loads only those specific segments into memory. This selective pre-loading maintains localization speed by having required data ready while avoiding the memory overhead of loading all segments.
3Quantity of substance
If sub-map files are loaded on-demand during localization cycles, then memory usage is optimized, but productivity decreases due to loading delays affecting multiple localization cycles
Solution Approach 1:
The system performs preliminary prediction of the set of group numbers to be loaded based on the current pose and desired trajectory, and loads the corresponding sub-map files into memory before they are needed for localization. This eliminates the loading wait time that would otherwise delay multiple localization cycles, thereby maintaining high productivity while optimizing memory usage by loading only necessary files.
Solution Approach 2:
The system uses feedback from the current pose and desired trajectory to continuously update the prediction of which sub-map files to load. This feedback mechanism ensures that the pre-loaded files are exactly those needed for upcoming localization operations, maximizing productivity without wasting memory on unnecessary data.
Data Source
AI summary
According to some aspects of the present disclosure, a method for loading a visual localization map is provided. The method may include: localizing a current pose; predicting, based on the current pose, a set of group numbers to be loaded for the visual localization map, wherein each group number in the set of group numbers to be loaded corresponds to a sub-map file of the visual localization map, wherein the visual localization map includes a master map file and a plurality of sub-map files, wherein the plurality of sub-map files respectively store map data of corresponding groups obtained by grouping the visual localization map based on key frames, and wherein key frame index information for indexing the plurality of sub-map files is stored in the master map file; and loading corresponding sub-map files based on the group numbers in the set of group numbers to be loaded.