Vehicle Obstacle Height Detection Using 2D Reference Line Projection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for determining the height of obstacles outside a vehicle, such as using light detection and ranging (LIDAR) or deep learning from 2D images, are either costly or inaccurate in providing exact height information.
Innovation Solution
An apparatus and method utilizing a camera and processor to detect obstacles by projecting a reference line onto a 2D image, determining a reference lower point, and calculating the obstacle's height based on threshold values, without the need for expensive equipment or additional deep learning for height determination.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If LIDAR is used to determine the height of the obstacle, then the measurement precision is improved, but the device complexity and cost increase significantly
Solution Approach 1:
The patent uses a 2D image (a copy or projection of the 3D scene) to determine obstacle height. Instead of directly measuring 3D space with complex LIDAR, the system captures a 2D image and uses image processing techniques to infer height information, thereby reducing device complexity while maintaining measurement capability
Solution Approach 2:
The patent replaces the mechanical/optical measurement system (LIDAR) with an imaging and image processing system. By substituting direct physical measurement with optical imaging followed by computational analysis, the system achieves comparable height measurement precision with simpler and less expensive equipment
2Device complexity
If deep learning is performed on a 2D image to determine object height, then the device complexity is reduced, but the measurement precision deteriorates
Solution Approach 1:
The patent introduces a reference line as an additional dimensional element in the 2D image space. By projecting a reference line that corresponds to a known physical dimension (such as the height of a standard object or a calibrated reference), the system creates a spatial reference framework that enables accurate height measurement without requiring complex deep learning models
Solution Approach 2:
The reference line acts as an intermediary element between the 2D image and the 3D height measurement. This reference line provides a known spatial relationship that mediates the conversion from 2D image coordinates to 3D height information, enabling precise measurement through simple geometric calculations rather than complex neural network inference
3Measurement precision
If additional deep learning is performed for height determination beyond classification, then the measurement precision is improved, but the loss of time increases
Solution Approach 1:
The patent extracts only the necessary information (obstacle height) directly from the 2D image using simple geometric relationships and reference lines, rather than performing comprehensive deep learning analysis. By extracting only the specific height information needed through efficient image processing, the system avoids the time-consuming nature of full deep learning pipelines while maintaining measurement accuracy
Data Source
AI summary
An apparatus for determining a height of an abject includes a camera to acquire a two-dimensional (2D) image and a processor. The processor detects a target object corresponding to an obstacle from the 2D image. The processor also determines a reference lower point from among pixels positioned at a lower portion of the target object. The processor further projects a preset reference line to the 2D image such that a reference point of the preset reference line is matched with the reference lower point. The processor additionally determines a height of the target object based on at least one threshold value for marking a preset distance, from the reference point, on the reference line.