Vehicle Camera Dirt Detection Using Center-Peripheral Edge Strength
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vehicle monitoring devices inaccurately determine the presence and removal of dirt on camera lenses due to variations in edge strength across the camera image, particularly in wider view angles, leading to inappropriate support control decisions.
Innovation Solution
A monitor device that differentiates between center and peripheral areas of the camera image to accurately assess dirt presence by using threshold values for edge strength calculations, determining the entire dirt state based on edge strength in the center area and considering conditions for the outer area, thereby improving dirt detection accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If the camera has a wider view angle to photograph more area around the own vehicle, then the coverage area is improved, but the edge strength calculation accuracy in peripheral areas deteriorates
Solution Approach 1:
The camera image is divided into a center area and an outer area. The determination unit calculates edge strength for pixels in both areas separately. By segmenting the image, the system can apply different evaluation criteria to areas with different calculation accuracy characteristics, thereby resolving the contradiction between wide coverage and accurate measurement.
2Area of stationary object
If the device uses edge strength of pixels in the peripheral area for dirt detection, then the detection coverage is improved, but the determination accuracy deteriorates due to lower edge strength calculation accuracy
Solution Approach 1:
The determination unit applies different quality standards to different areas: it primarily relies on edge strength from the center area where calculation is accurate, while considering the outer area only as supplementary. This local quality approach ensures high determination accuracy by weighting areas according to their reliability.
Solution Approach 2:
Instead of using peripheral area data as the primary basis for determination (which would expand coverage but reduce accuracy), the system inverts the approach by using center area data as primary and outer area data as secondary confirmation, thereby prioritizing accuracy over coverage in the determination process.
3Device complexity
If the device determines dirt presence based on edge strength without considering camera image properties, then the determination process is simplified, but the determination accuracy deteriorates
Solution Approach 1:
The determination unit changes the parameter selection based on the spatial location within the image. It uses edge strength values from the center area as the primary parameter for determination, while using outer area edge strength as a secondary parameter. This parameter change strategy maintains relatively simple processing while significantly improving accuracy by adapting to camera image properties.
Data Source
AI summary
A monitor ECU 10 performs a support control based on a camera image photographed through a protection window by a camera 21. The ECU determines that a protection window state is an entire dirt state, when a first area index value calculated based on edge strength of pixels in a first area including a center of the camera image is smaller than a threshold dirt value. The ECU determines that the entire dirt state ends, when at least one of a first condition or a second condition is established. The first condition is established when the first area index value is equal to or greater than a first threshold end value. The second condition is established when a second area index value calculated based on the edge strength in a second area except the first area is equal to or greater than a second threshold end value.