Environment Recognition Device Luminance Correction for Traffic Light Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing environment recognition systems face challenges in accurately identifying target objects due to the adverse effects of strong environment light, such as sunlight, especially when the windshield is fogged, which can alter the color phase of light sources like traffic lights, leading to decreased accuracy in object specification.
Innovation Solution
An environment recognition device and method that include a data retaining unit for luminance associations, a luminance obtaining unit, a white balance deriving unit, a corrected luminance deriving unit, and a specific object determining unit, which corrects luminance values based on white balance and color correction intensity to improve object identification accuracy, even under the influence of environment light.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If white balance correction is performed on the obtained luminance to adapt to environment light, then the image can be adjusted to normal color conditions, but the color phase of light sources such as traffic lights may be altered, deteriorating the accuracy of specifying the target object
Solution Approach 1:
The patent segments the luminance correction process into two distinct paths: one for general environment light adaptation (white balance correction) and another for preserving light source color phases. By separating these functions, the system can apply white balance to the overall image while protecting traffic light colors from being altered, thus resolving the contradiction between adaptability and measurement precision.
Solution Approach 2:
The patent applies different quality treatments to different regions of the image: white balance correction is applied to general areas to adapt to environment light, while light source regions (traffic lights) are excluded from this correction to preserve their original color phases. This local differentiation allows simultaneous achievement of environment adaptability and target specification accuracy.
2Illumination intensity
If the entire captured image is corrected for white balance under strong environment light, then the overall image color can be normalized, but the color phase of specific light sources changes, leading to decreased identification accuracy
Solution Approach 1:
The patent extracts and identifies light source regions (traffic lights) from the captured image before applying white balance correction. By taking out these critical regions and excluding them from the correction process, the system can normalize the overall image brightness while preserving the original color phases of light sources, thus maintaining identification accuracy.
Solution Approach 2:
The patent introduces an intermediary step of light source detection and masking between image capture and white balance correction. This intermediary mechanism allows the system to selectively protect light source regions from correction while allowing general areas to be normalized, resolving the conflict between overall illumination normalization and light source color preservation.
Data Source
AI summary
There are provided an environment, recognition device and an environment recognition method. The environment, recognition device obtains a luminance of a target portion in a detection area; obtains a height of the target portion; derives a white balance correction value, assuming that white balancing is performed to the obtained luminance; derives the corrected luminance by subtracting the white balance correction value and a color correction value based upon a color correction intensity indicating a degree of an influence of environment light from the obtained luminance; and provisionally determines a specific object corresponding to the target portion from the corrected luminance of the target portion based on an association of a luminance range and the specific object retained in a data retaining unit.