Radar-Camera Fusion Tracking Using Stationary Object Speed Reference
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing object detection systems using radar and camera fusion fail to accurately detect objects due to the lack of consideration for the vehicle's speed, leading to reduced radar performance in tracking objects.
Innovation Solution
An object detection apparatus that combines radar and camera data to determine stationary objects by comparing radar and camera positions and types, calculates the vehicle's speed based on stationary object speeds, and uses this information to enhance tracking accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If radar detection is performed without using own vehicle speed, then the detection system can operate with simpler processing, but the tracking performance of the radar is lowered and object detection accuracy deteriorates
Solution Approach 1:
The system performs preliminary identification of stationary objects using camera image processing before radar tracking. By pre-identifying which objects are stationary through image recognition, the system prepares the necessary information in advance that enables the radar to use own vehicle speed for tracking moving objects, thereby improving detection accuracy without excessive complexity
Solution Approach 2:
The detection system is divided into separate functional modules: camera image processing unit for identifying stationary objects, radar signal processing unit for detecting object positions and speeds, and fusion processing unit for integrating the data. This segmentation allows each module to perform its specific function efficiently, maintaining reliability while managing processing complexity through modular architecture
2Measurement precision
If the system processes all detected objects as moving objects, then no additional classification processing is needed, but the tracking accuracy deteriorates because stationary objects are incorrectly tracked as moving objects
Solution Approach 1:
The camera image processing unit acts as an intermediary that provides stationary object identification information to the radar signal processing unit. This intermediary component enables the radar to distinguish between stationary and moving objects without requiring the radar itself to perform complex classification, thereby improving tracking accuracy while managing complexity through specialized subsystems
Solution Approach 2:
The system replaces mechanical/radar-based object classification with optical/image-based classification performed by the camera system. By using image processing to identify stationary objects, the system substitutes a different sensing modality (optical) for the traditional radar-based classification approach, improving measurement precision while leveraging the strengths of each sensing type
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The apparatus accurately distinguishes between moving and stationary objects, improving tracking performance and overall detection accuracy by integrating vehicle speed calculations.
Implementation Method 1
a radar that emits electromagnetic waves and receives reflected signals from objects
Implementation Method 2
a camera that acquires image data on the objects by imaging the objects
Data Source
AI summary
An object detection apparatus includes: a radar signal processor that detects radar detection positions and radar detection speeds of objects; a camera image processor that detects camera detection positions and object types of some of the objects; and a fusion processor that outputs the radar detection position, the radar detection speed, and the object type of the object to a vehicle control device, a position of the object indicated by the radar detection position being identical to a position of the object indicated by the camera detection position, wherein the radar signal processor determines the stationary object based on the radar detection positions, the camera detection position of the moving object, and the object type of the moving object, calculates an own vehicle speed based on the radar detection speed of the stationary object, and causes the radar to track the objects based on the own vehicle speed.