Vehicle Camera Fog Detection via Spectral Analysis

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current fog detection methods in vehicles often rely on active lighting, distance determination, or specific sensors, and struggle to accurately differentiate between fog and other weather conditions like snow, leading to incorrect detections.

Innovation Solution

A method that evaluates the spectral properties of aerosols using imaging techniques without active lighting, by analyzing the absorption and dispersion of light through aerosols, which are identified through color filtering and parameter comparison in images, allowing for precise differentiation and density determination of aerosols.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If active lighting elements (e.g., infrared LED) are used for fog detection, then detection capability is improved, but device complexity and energy consumption increase

Engineering Contradiction:
Improvefog detection capabilityVSAvoiddevice complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The system uses the vehicle's existing spotlight/headlight illumination and standard color camera to perform fog detection, rather than adding dedicated active sensing elements. The fog detection function is achieved by evaluating spectral properties of ambient light reflected from or transmitted through aerosols using the existing imaging system.

Inventive Principle:
Principle #25Self-service

2Measurement precision

If distance determination or visibility hypothesis is used in imaging techniques, then fog detection accuracy is improved, but device complexity increases

Engineering Contradiction:
Improvefog detection accuracyVSAvoiddevice complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The invention extracts and evaluates specific spectral information (color ratios, particularly red channel intensity relative to other channels) from standard color images to detect fog, separating the fog detection function from distance measurement or visibility hypothesis requirements. This allows fog detection using only the existing color camera without additional ranging sensors.

Inventive Principle:
Principle #2Taking out (Extraction)

3Device complexity

If grayscale distribution in image areas is evaluated for fog detection, then detection method is simplified, but false detection rate increases due to inability to differentiate from other weather conditions

Engineering Contradiction:
Improvedetection method simplicityVSAvoiddetection accuracy
Core Design Contradiction:
Device complexityVSReliability

Solution Approach 1:

The system uses color information from the color camera, specifically evaluating the intensity ratio in the red color channel relative to other channels. Fog particles scatter and absorb light in wavelength-dependent ways that produce characteristic color signatures, allowing differentiation from snow, rain, or other weather conditions that may have similar grayscale characteristics.

Inventive Principle:
Principle #32Color changes

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

This approach enables accurate fog detection without active lighting or special sensors, reduces misinterpretation of other weather conditions, and can be implemented using existing vehicle cameras, thereby enhancing safety and comfort by optimizing lighting and driver assistance systems.

Implementation Method 1

The underlying physical effect is the increased absorption of light when passing through aerosols, where the light rays are absorbed or refracted by the particles or molecules

Methodology Applied
Scientific EffectAbsorption (EM radiation): Absorption (EM radiation)

Implementation Method 2

This absorption and dispersion are wavelength-dependent

Methodology Applied
Scientific EffectDispersion (of waves): Dispersion (of waves)

Implementation Method 3

either in the form of a color image or in the form of a grayscale image

Methodology Applied
Scientific EffectPhotoelectric Effect: Photoelectric Effect

Data Source

PatentEP2542881B1Method and apparatus for detecting mist by means of spectroscopy
Publication Date: 2024.03.13 ROBERT BOSCH GMBH
  • EP2542881B1 patent drawingFigure 1~2
  • EP2542881B1 patent drawingFigure 3a~3f
  • EP2542881B1 patent drawingFigure 4

AI summary

The invention relates to the detection or determination of density or the classification of an aerosol by means of spectroscopy. Using an image recorded by a camera for a vehicle, a first value of a parameter is determined from the at least one image in a first step with first color filtering, and a second value for the same parameter is determined in a second step without any color filtering or with second color filtering that is different from the preceding color filtering. The values determined in the at least two steps are compared and the detection or determination of density or the classification of the aerosol is carried out on the basis of the result of the comparison.