18140246. METHOD AND SYSTEM FOR DETECTING OBJECT simplified abstract (KIA CORPORATION)
METHOD AND SYSTEM FOR DETECTING OBJECT
Organization Name
Inventor(s)
METHOD AND SYSTEM FOR DETECTING OBJECT - A simplified explanation of the abstract
This abstract first appeared for US patent application 18140246 titled 'METHOD AND SYSTEM FOR DETECTING OBJECT
Simplified Explanation
The present disclosure describes a method and system for detecting an object in a region of interest of a vehicle using LiDAR sensor data.
- Determining object's state (turning or straight moving)
- Locating a reference point for the object
- Extracting outline points from LiDAR data
- Generating virtual points based on outline points
- Creating a bounding box for the object based on outline points and virtual points
Potential Applications
This technology can be applied in autonomous vehicles, collision avoidance systems, and object detection in various industries such as transportation and robotics.
Problems Solved
- Improved object detection accuracy - Efficient processing of LiDAR sensor data - Enhanced safety in vehicle navigation
Benefits
- Increased safety on the road - Enhanced object detection capabilities - Improved decision-making for autonomous systems
Potential Commercial Applications
- Automotive industry for autonomous driving - Robotics industry for object detection and navigation systems
Possible Prior Art
One possible prior art in this field is the use of LiDAR sensors for object detection in autonomous vehicles. Previous methods may have focused on specific aspects of object detection, while the present disclosure provides a comprehensive approach to determining object state and generating accurate bounding boxes.
Unanswered Questions
How does this technology handle complex road scenarios with multiple objects?
The abstract does not specify how the method deals with situations where there are multiple objects in the region of interest. It would be interesting to know if the system can differentiate between different objects and track them individually.
What is the computational cost of implementing this method in real-time applications?
The abstract does not mention the computational resources required to perform the object detection process in real-time. Understanding the computational efficiency of the system would be crucial for assessing its practicality in various applications.
Original Abstract Submitted
The present disclosure relates to a method and a system for detecting an object. A method of detecting an object, according to an embodiment of the present disclosure, includes determining whether the object in a region of interest of a (host) vehicle is in a turning state or a straight moving state, determining a location of a reference point for the object based on the determination, extracting first outline points from LiDAR points for the object, which are included in a point cloud received from a LiDAR sensor, based on the reference point, generating at least one virtual point based on a length of a line segment connecting the first outline points, and generating a bounding box of the object corresponding to a heading direction of the object based on second outline points including the first outline points and the at least one virtual point.
