3D Object Detection Device Shadow Differentiation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing three-dimensional object detection systems face challenges in accurately differentiating the shadows of trees with periodicity and irregularity from other vehicles traveling in adjacent lanes, leading to erroneous detections.
Innovation Solution
A three-dimensional object detection device that calculates periodicity and irregularity evaluation values based on differential waveform or edge information to distinguish shadows of trees from other vehicles, using threshold values to determine the nature of detected objects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If pattern matching is used to detect three-dimensional objects, then detection capability is provided, but shadows of trees with periodicity and irregularity are erroneously detected as vehicles
Solution Approach 1:
The invention introduces two new evaluation parameters (periodicity evaluation value and irregularity evaluation value) to characterize detected objects. By calculating these parameters from differential waveform information and comparing them against threshold values, the system can distinguish between tree shadows (which exhibit specific periodicity and irregularity patterns) and actual vehicles, thereby eliminating erroneous detections while maintaining reliable detection capability
Solution Approach 2:
The detection process is segmented into multiple evaluation stages: first calculating differential waveform information from captured images, then computing periodicity and irregularity evaluation values separately, and finally combining these evaluations with threshold comparisons to make the final detection determination. This segmented approach allows each aspect of object characterization to be analyzed independently, improving overall detection accuracy
2Measurement precision
If simple detection methods are used, then detection speed is maintained, but precision in differentiating tree shadows from vehicles is insufficient
Solution Approach 1:
The invention replaces complex mechanical or hardware-based differentiation methods with computational analysis of image differential waveforms. By substituting physical complexity with mathematical processing of periodicity and irregularity patterns, the system achieves high precision in differentiating tree shadows from vehicles without requiring additional complex detection hardware or mechanisms
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Provided are a three-dimensional object detection unit (33, 37) for detecting a three-dimensional object present in detection areas (A1, A2) rearward of a vehicle based on image information rearward of the vehicle from a camera (10); a stationary object assessment unit (38) for calculating a periodicity evaluation value and an irregularity evaluation value, and determining that a detected three-dimensional object is a shadow of a tree (tree shadows (Q1)) present along the road traveled by the host vehicle (V) when the periodicity evaluation value is equal to or greater than a first periodicity evaluation threshold value and less than a second periodicity evaluation threshold value, and the irregularity evaluation value is less than an irregularity evaluation threshold value; and a control unit (39) for controlling various processing, wherein the control unit (39) suppresses determination of a detected three-dimensional object as another vehicle (VX) when the detected three-dimensional object was determined by the stationary object assessment unit (38) to be tree shadows (Q1).