Camera-Based Object Projection for Autonomous Vehicle Navigation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
LIDAR sensors have limited range, resulting in incomplete perception of the surrounding environment, which restricts the ability of autonomous vehicles to plan maneuvers and navigate effectively, as they cannot detect objects beyond their range until they are within scanning distance.
Innovation Solution
An object projection system that uses camera images to identify and project nearby objects into a map, determining their relative location and updating the obstacle map, allowing for enhanced environmental perception and navigation planning even beyond the range of LIDAR sensors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If LIDAR sensor is used to detect objects, then measurement precision is improved, but the detection range is limited
Solution Approach 1:
The patent uses camera images as an intermediary to extend the detection range of the LIDAR sensor. The camera captures visual information of distant objects, which then guides the LIDAR sensor to focus on those specific objects, effectively combining the long-range capability of cameras with the precision measurement capability of LIDAR.
Solution Approach 2:
The patent transitions from relying solely on LIDAR's direct measurement capability to a multi-dimensional approach that combines camera visual detection with LIDAR precision measurement. This dimensional shift allows the system to first identify objects visually at any distance, then apply precise LIDAR measurement only when needed.
2Loss of information
If LIDAR sensor range is increased, then perception of surrounding environment is improved, but device complexity increases
Solution Approach 1:
The patent applies partial action by using the LIDAR sensor only when necessary - specifically when a camera identifies a potential object of interest. Instead of continuously operating LIDAR at full power across all ranges, the system selectively activates LIDAR for specific targets, reducing overall system complexity while maintaining complete environmental perception.
Solution Approach 2:
The camera performs preliminary detection of objects in the environment before the LIDAR sensor is activated. This preliminary action allows the system to identify potential objects of interest at a distance, so that when LIDAR is subsequently activated, it can focus its measurement capabilities on already-identified targets, thereby reducing the need for complex high-range LIDAR configurations.
3Reliability
If LIDAR sensor operates at maximum range, then detection capability is improved, but energy consumption increases
Solution Approach 1:
The system uses partial action by activating the LIDAR sensor only when and where needed based on camera detection results. Instead of operating LIDAR at maximum range continuously to ensure reliable detection, the system selectively engages LIDAR for specific objects identified by the camera, thereby maintaining detection reliability for critical objects while significantly reducing overall energy consumption.
Solution Approach 2:
The camera serves the LIDAR sensor by pre-identifying objects of interest, allowing the LIDAR to focus its energy on specific targets rather than scanning the entire environment at maximum range. This self-service relationship between the camera and LIDAR enables reliable object detection while minimizing energy consumption through targeted rather than exhaustive scanning.
Data Source
AI summary
System, methods, and other embodiments described herein relate to locating an object within an image and projecting the object into a map. In one embodiment, a method includes, responsive to identifying a nearby vehicle in an image of an external environment surrounding a scanning vehicle, determining a relative location of the nearby vehicle in relation to the scanning vehicle using the image. The method includes projecting the nearby vehicle into a map according to at least the relative location determined from the image. The method includes controlling the scanning vehicle according to the map.