18113785. SYSTEMS AND METHODS FOR DETECTING OBJECTS BASED ON LIDAR DATA simplified abstract (DENSO CORPORATION)
Contents
SYSTEMS AND METHODS FOR DETECTING OBJECTS BASED ON LIDAR DATA
Organization Name
Inventor(s)
SYSTEMS AND METHODS FOR DETECTING OBJECTS BASED ON LIDAR DATA - A simplified explanation of the abstract
This abstract first appeared for US patent application 18113785 titled 'SYSTEMS AND METHODS FOR DETECTING OBJECTS BASED ON LIDAR DATA
The method described in the patent application involves using lidar data to generate multiple inputs, running a convolutional neural network routine on each input to create bounding boxes, dimensional characteristics, and confidence scores for objects detected in the lidar data. The routine then selects target bounding boxes based on confidence scores to identify objects.
- Lidar data is used to create multiple inputs.
- A convolutional neural network routine is applied to each input to generate bounding boxes.
- The routine determines dimensional characteristics and confidence scores for each bounding box.
- Target bounding boxes are selected based on confidence scores to identify objects.
Potential Applications: - Autonomous vehicles for object detection and collision avoidance. - Surveillance systems for tracking and identifying objects in real-time.
Problems Solved: - Efficient and accurate object detection in lidar data. - Automated selection of target objects based on confidence scores.
Benefits: - Improved safety in autonomous vehicles. - Enhanced surveillance capabilities for security systems.
Commercial Applications: Title: Advanced Object Detection System for Autonomous Vehicles This technology can be utilized in the automotive industry for developing advanced driver-assistance systems (ADAS) and self-driving vehicles. It can also be integrated into surveillance systems for enhanced security measures.
Prior Art: Prior research in lidar-based object detection systems and convolutional neural networks for image processing can provide insights into similar technologies.
Frequently Updated Research: Ongoing research in lidar technology and neural network algorithms may lead to advancements in object detection systems for various applications.
Questions about Lidar-based Object Detection System: 1. How does lidar technology improve object detection compared to other sensing technologies? 2. What are the limitations of using convolutional neural networks for object detection in lidar data?
Original Abstract Submitted
A method includes generating a plurality of lidar inputs based on the lidar data and performing, for each lidar input from among the plurality of lidar inputs, a convolutional neural network (CNN) routine to: generate a plurality of bounding boxes based on the lidar inputs, determine one or more dimensional characteristics associated with each of the plurality of bounding boxes, and generate a confidence score associated with each of the plurality of bounding boxes. The method includes selecting one or more target bounding boxes from among the plurality of bounding boxes based on the confidence score associated with each of the plurality of bounding boxes and identifying the one or more objects based on the one or more target bounding boxes.
