Image-Based Visibility Measurement Using Dark Channel Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing visibility detection systems, such as laser-based meters, are expensive and impractical for widespread deployment, and vehicles without advanced sensors or connectivity face challenges in addressing low visibility conditions, leading to safety concerns.
Innovation Solution
An edge device equipped with an image sensor and processing capabilities determines visibility measures by analyzing images to calculate a dark channel, transmission map, and visual contrast, generating real-time data without requiring communication with a back-end server.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If laser-based visibility meters are deployed, then visibility detection accuracy is improved, but deployment cost and complexity increase significantly
Solution Approach 1:
The patent uses standard camera images as a copy or alternative representation of the visual environment, replacing the need for specialized laser-based visibility meters. By processing regular camera feeds through computational algorithms (dark channel prior, transmission map calculation), the system derives visibility information without requiring expensive dedicated sensing hardware, thus achieving cost-effective deployment while maintaining reasonable measurement accuracy
Solution Approach 2:
The patent substitutes complex optical measurement systems (laser-based meters) with computational image processing methods. Instead of using physical laser ranging and atmospheric scattering measurements, the system uses algorithmic analysis of standard camera images to infer visibility conditions, replacing mechanical/optical complexity with computational simplicity
2Reliability
If advanced sensors are installed in vehicles, then real-time visibility data is obtained, but vehicle cost and complexity increase
Solution Approach 1:
The patent enables standard cameras to perform multiple functions - both general imaging and specialized visibility measurement. By processing the same camera feed for both navigation/imaging purposes and visibility assessment, the system eliminates the need for dedicated visibility sensors in vehicles, reducing complexity while maintaining reliable real-time visibility data through infrastructure-based processing
Solution Approach 2:
The system allows standard vehicle cameras to serve themselves for visibility measurement without requiring additional specialized sensors. The infrastructure processing unit analyzes the regular camera feeds that vehicles already capture for other purposes, enabling vehicles to obtain visibility data using their existing imaging capabilities
3Adaptability or versatility
If infrastructure-based visibility monitoring is implemented, then widespread deployment is achieved, but measurement precision may be reduced compared to specialized sensors
Solution Approach 1:
The patent transforms the measurement approach by changing from direct physical measurement (laser ranging) to computational inference from image intensity distributions. By applying dark channel prior theory and transmission map calculations to standard camera images, the system achieves reasonable visibility measurement accuracy using infrastructure-based cameras, balancing scalability with acceptable precision for traffic safety applications
Data Source
AI summary
An edge device can include a processing device, an image sensor, and a memory having instructions that are executable by the processing device for causing the processing device to perform operations. The processing device can receive, from the image sensor, an image of an environment. The processing device can determine a visibility measure corresponding to the environment by determining a dark channel of the image, determining, based on the dark channel of the image, a transmission map of the image, and determining, based on the transmission map, a visual contrast of the image. The processing device can generate information corresponding to the visibility measure.