Vehicle Fallen-Object Detection Using Headlight Shadows
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies struggle to accurately detect low-height obstacles or fallen objects on a road surface in front of a vehicle using only the image of the obstacle, leading to detection difficulties.
Innovation Solution
A surrounding situation recognition device that utilizes a camera positioned below the headlight and detects obstacles by identifying shadows created on the road surface by the obstacles, which are positioned on the opposite side of the headlight, using a model trained on teacher data to enhance detection accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If only the portion corresponding to the fallen object in the image is used for detection, then the detection method is simple, but low-height obstacles become difficult to detect
Solution Approach 1:
The patent transitions from detecting obstacles only in the vertical dimension (obstacle height) to detecting obstacles in the horizontal dimension by utilizing shadow information. The shadow extends horizontally from the obstacle away from the light source, providing a detection dimension that works effectively for low-height obstacles where vertical detection fails.
Solution Approach 2:
The shadow acts as an intermediary element between the obstacle and the camera. Instead of directly detecting the obstacle itself (which is difficult for low-height objects), the system detects the shadow cast by the obstacle, which provides sufficient information to infer the obstacle's presence and location.
2Measurement precision
If the camera is positioned below the headlight to capture shadow information, then low-height obstacle detection improves, but the camera positioning becomes more complex
Solution Approach 1:
The camera is positioned to serve multiple functions: capturing images for obstacle detection and capturing shadow information for low-height obstacle detection. By using a single camera position that can detect both the obstacle and its shadow, the system avoids the need for separate positioning systems for different detection tasks.
Solution Approach 2:
The existing camera system is utilized to perform the additional function of shadow detection without requiring separate dedicated sensors or complex positioning mechanisms. The camera's natural field of view and imaging capability are leveraged to detect shadows, making the system self-sufficient.
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
Enables accurate detection of low-height obstacles by leveraging shadows, improving detection capabilities beyond conventional methods that focus solely on the obstacles themselves.
Implementation Method 1
light emitted from the headlight
Implementation Method 2
a shadow created on the road surface by the fallen object blocking light emitted from the headlight
Data Source
AI summary
A surrounding situation recognition device acquires an image of the front of a vehicle shot by a camera located in a position out of a lower portion which is lower than a line segment connecting a headlight of the vehicle and a fallen object on a road surface in front of the vehicle, the line segment being included in a vertical plane and the lower portion being included in the vertical plane, and detects the fallen object included in the image by using the fact that a shadow created on the road surface by the fallen object blocking light emitted from the headlight is included in the image, the shadow being positioned on the opposite side of the headlight across the fallen object.