Multi-Frame Object Segmentation for Sparse Vehicle Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face challenges in processing vast volumes of data from various sources, including captured image data, GPS data, and sensor data, which can limit navigation accuracy and efficiency, especially in updating and storing traditional maps.
Innovation Solution
A system and method for identifying objects in a vehicle's environment by assigning pixel descriptor values to images, determining object descriptors, and comparing them across multiple images to recognize common objects, using a combination of cameras and processing units for navigation, and constructing a sparse map for efficient data management.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional mapping technology is used to navigate, then navigation coverage is comprehensive, but data storage and update requirements become overwhelming
Solution Approach 1:
The patent segments the environment into discrete objects (vehicles, pedestrians, signs, etc.) and represents them individually in the map data structure, rather than storing continuous comprehensive map data. This segmentation allows the system to store only relevant object information needed for navigation decisions.
Solution Approach 2:
The system extracts only the essential object representations needed for navigation from the complete environmental data, storing merely the critical elements (object descriptors, positions, types) rather than the full voluminous map data required by traditional mapping technology.
2Measurement precision
If vast volumes of image data are processed to identify objects, then object recognition accuracy improves, but data processing complexity and time increase
Solution Approach 1:
The system performs preliminary actions by assigning pixel descriptor values to images and determining object descriptors before comparison. This preprocessing organizes the data in advance, making subsequent object identification and matching across frames more efficient and less complex.
Solution Approach 2:
The patent creates simplified copies of object information through descriptors that represent complex visual data in a condensed format. These object descriptors serve as efficient representations that capture essential object characteristics without requiring processing of the full original image data.
3Reliability
If multiple image frames are analyzed to track objects, then object tracking reliability improves, but processing time and computational load increase
Solution Approach 1:
The system uses object descriptors as simplified copies that represent objects across multiple frames. By comparing these compact descriptor representations rather than full images, the system achieves reliable multi-frame object tracking with significantly reduced processing time and computational requirements.
Data Source
AI summary
Systems and methods for identifying objects in an environment of a host vehicle are disclosed. In one implementation, a system includes a processor configured to receive images representative of the environment of the host vehicle; assign first pixel descriptor values to a plurality of pixels associated with a first image and second pixel descriptor values to a plurality of pixels associated with a second image; identify object representations in the first image and the second image based on at the first pixel descriptor values and the second pixel descriptor values, respectively; determine a first object descriptor and a second object descriptor based on the first pixel descriptor values and the second pixel descriptor values, respectively; and based on a comparison of the first object descriptor and the second object descriptor, output an indication that the object representations in the first image and the second image represent a common object.