3D Object Detection Device Adapting Edge Extraction to Light Environment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional three-dimensional object detection systems struggle to accurately detect objects near a host vehicle when the camera lens is clouded by foreign matter, such as water stains, leading to blurred images due to varying light environments between daytime and nighttime.
Innovation Solution
A three-dimensional object detection device that extracts edges from specific areas, including a horizon reference area during the day and road edge reference areas at night, calculates edge intensity, and adjusts detection parameters based on these intensities to compensate for lens clouding and light environment changes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If edge extraction is performed in a fixed area regardless of light environment, then the device complexity is reduced, but the detection precision deteriorates when the lens is clouded
Solution Approach 1:
The patent applies dynamics by making the edge extraction area variable rather than fixed. The system dynamically selects between a first edge extraction area (including horizon reference area) during daytime and a second edge extraction area (including road edge reference area) during nighttime based on light environment detection. This dynamic adaptation allows the system to maintain high detection precision under varying conditions without requiring multiple permanent configurations.
Solution Approach 2:
The patent changes the parameter of edge extraction area selection based on light environment conditions. By detecting whether it is daytime or nighttime and accordingly selecting different extraction areas and calculating different edge intensities (first edge intensity or second edge intensity), the system adapts to varying light conditions. This parameter change enables accurate lens clouding detection regardless of the time of day.
2Reliability
If a single edge extraction area is used for both daytime and nighttime, then the ease of operation is improved, but the reliability of detection deteriorates under varying light conditions
Solution Approach 1:
The system dynamically adjusts the edge extraction area based on detected light environment conditions. During daytime, it uses a first edge extraction area that includes a horizon reference area, while during nighttime, it switches to a second edge extraction area that includes a road edge reference area. This dynamic switching ensures reliable detection across different operating conditions without requiring manual intervention.
Solution Approach 2:
The patent changes operational parameters (edge extraction area and edge intensity calculation method) according to light environment conditions. By automatically detecting whether it is daytime or nighttime and相应地 changing the extraction area parameter, the system maintains high reliability without complicating the user interface or operation.
3Measurement precision
If edge intensity calculation is performed without considering light environment, then the productivity is improved, but the measurement precision deteriorates when lens clouding occurs
Solution Approach 1:
The patent changes the calculation parameters for edge intensity based on light environment. The system calculates a first edge intensity using a first edge extraction area during daytime and a second edge intensity using a second edge extraction area during nighttime. This parameter adaptation ensures accurate lens clouding detection while maintaining efficient processing through automated selection based on time of day.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A three-dimensional object detection device wherein, an assessment is made regarding whether the present is daytime or nighttime when detecting a three-dimensional object based on a captured image; when an assessment has been made that the present is daytime, the edges of a subject are extracted from a first edge extraction area, including a horizon reference area that references the horizon, and a threshold value for detecting the three-dimensional object is set based on the intensity of the edges in the first edge extraction area; when an assessment has been made that the present is nighttime, the edges of a subject are extracted from a second edge extraction area, including a road edge reference area that references the road edge, and a threshold value for detecting the three-dimensional object is set based on the intensity of the edges that are extracted from the second edge extraction area.