20240037961. SYSTEMS AND METHODS FOR DETECTING LANES USING A SEGMENTED IMAGE AND SEMANTIC CONTEXT simplified abstract (Toyota Research Institute, Inc.)
SYSTEMS AND METHODS FOR DETECTING LANES USING A SEGMENTED IMAGE AND SEMANTIC CONTEXT
Organization Name
Toyota Research Institute, Inc.
Inventor(s)
Shunsho Kaku of Mountain View CA (US)
Jeffrey M. Walls of Ann Arbor MI (US)
Kun-Hsin Chen of Mountain View CA (US)
Steven A. Parkison of Ann Arbor MI (US)
SYSTEMS AND METHODS FOR DETECTING LANES USING A SEGMENTED IMAGE AND SEMANTIC CONTEXT - A simplified explanation of the abstract
This abstract first appeared for US patent application 20240037961 titled 'SYSTEMS AND METHODS FOR DETECTING LANES USING A SEGMENTED IMAGE AND SEMANTIC CONTEXT
Simplified Explanation
The patent application describes a system and method for detecting lanes in a driving scene by segmenting road regions using an ontology to derive semantic context. Here are the key points:
- The method involves segmenting an image of a driving scene into lane lines and road regions using an ontology, without relying on maps.
- A pixel subset from the image contains lane information derived from the ontology.
- Pixel depth for the lane lines and road regions is computed using a model.
- 3D context is derived by establishing relations between the semantics and pixel depth.
- These relations infer a driving lane for a vehicle based on the types of lane lines and road regions adjacent to it.
- The method also includes executing a task to control the vehicle on the driving lane using the 3D context.
Potential Applications:
- Autonomous driving systems: This technology can be used in autonomous vehicles to accurately detect and navigate within lanes, improving safety and efficiency.
- Advanced driver assistance systems (ADAS): The system can enhance existing ADAS by providing more accurate lane detection and assisting with lane keeping and lane departure warning.
- Traffic management: The technology can be applied in traffic management systems to monitor and analyze lane usage, helping optimize traffic flow and reduce congestion.
Problems Solved:
- Accurate lane detection: The system addresses the challenge of accurately detecting lanes in a driving scene, even in complex and dynamic environments.
- Semantic understanding: By using an ontology to derive semantic context, the system can understand the meaning and relationships of different lane and road region types, improving lane detection accuracy.
Benefits:
- Improved safety: Accurate lane detection and understanding of semantic context can help prevent accidents by enabling better lane keeping and warning systems.
- Enhanced efficiency: The technology can optimize traffic flow by providing real-time lane usage information, reducing congestion and improving overall driving experience.
- Adaptability: The system can work independently of maps, making it suitable for various driving scenarios and locations.
Original Abstract Submitted
system, methods, and other embodiments described herein relate to the detection of lanes in a driving scene through segmenting road regions using an ontology enhanced to derive semantic context. in one embodiment, a method includes segmenting an image of a driving scene, independent of maps, by lane lines and road regions defined by an ontology and a pixel subset from the image has semantics of lane information from the ontology. the method also includes computing pixel depth from the image for the lane lines and the road regions using a model. the method also includes deriving 3d context using relations between the semantics and the pixel depth, the relations infer a driving lane for a vehicle from types of the lanes lines and the road regions adjacent to the driving lane. the method also includes executing a task to control the vehicle on the driving lane using the 3d context.
