Histogram-Based Fog Detection in Network Cameras
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Network cameras struggle to maintain image visibility in foggy or misty conditions due to varying contrast, and existing methods require auxiliary lights or multiple image captures, leading to accuracy issues and long processing times.
Innovation Solution
An image processing apparatus that determines fog or mist presence by analyzing the brightness histogram of captured images, setting preliminary determination points and thresholds to assess pixel distribution, allowing for contrast emphasis without auxiliary lights or multiple captures.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If auxiliary light is applied to determine fog or mist presence, then visibility of image capturing conditions can be improved, but device complexity and processing time increase
Solution Approach 1:
The patent extracts the fog/mist detection function from the auxiliary light system and implements it using only the main light source and histogram analysis. By removing the auxiliary light requirement, the system achieves fog detection without additional device complexity while maintaining measurement precision through statistical analysis of pixel brightness distribution.
Solution Approach 2:
The system uses its own main light source and captured image data to perform fog detection, rather than requiring separate auxiliary lighting. The histogram analysis leverages the existing image information to self-determine fog presence, eliminating the need for external assistance and reducing processing time.
2Measurement precision
If multiple images are captured with and without auxiliary light, then fog or mist detection accuracy improves, but processing time increases
Solution Approach 1:
The patent performs preliminary histogram analysis on the captured image to detect fog presence before full image processing. By calculating the histogram and analyzing pixel brightness distribution in advance, the system can quickly determine fog conditions without requiring multiple image captures, thus reducing processing time while maintaining detection accuracy.
Solution Approach 2:
Instead of capturing multiple complete images for fog detection, the system performs partial analysis using histogram data from a single image. This partial action approach extracts sufficient information for fog detection without the time cost of multiple captures, achieving acceptable accuracy with reduced processing overhead.
3Measurement precision
If contrast emphasis is applied when fog or mist is present, then image visibility improves, but processing complexity increases
Solution Approach 1:
The patent changes the gamma parameter based on fog detection results to adjust contrast. When fog is detected through histogram analysis, the system applies gamma correction with adjusted parameters to enhance visibility. This parameter-based approach provides effective contrast control without requiring complex processing systems, as gamma correction is a standard and efficient image processing operation.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A histogram is detected from an acquired image, and a ratio of a pixel distribution for a range from a preliminarily set low-brightness-side determination point to a preliminarily set high-brightness-side determination point to the entire histogram is calculated for the detected histogram. In a case where the ratio exceeds a preliminarily set threshold, it is estimated that there is fog or mist and a subject looks hazy.