Vehicle Obstacle Range Detection via RGB-to-3D Point Clouds
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In autonomous driving, vehicles face challenges in accurately detecting obstacles, leading to unsafe driving conditions due to uncertainty in obstacle range, which affects user experience and safety.
Innovation Solution
A method utilizing a vehicle-mounted electronic device that captures RGB images, processes them using a trained depth estimation model to generate depth images, converts these into 3D point cloud maps, and determines 3D regions of interest to identify obstacles, triggering alarms or automatic braking as necessary.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional obstacle detection methods are used, then the detection process is simple, but the measurement precision of obstacle range is insufficient
Solution Approach 1:
The patent transforms 2D RGB images into 3D point cloud maps by introducing depth information through depth estimation models. This dimensional transformation enables accurate measurement of obstacle ranges in three-dimensional space, directly resolving the measurement precision issue while maintaining reasonable system complexity through software-based processing.
Solution Approach 2:
The patent introduces depth images as an intermediary between RGB images and obstacle detection. The depth estimation model generates depth images that provide distance information, which then serves as a bridge to create accurate 3D point cloud representations of obstacles, improving measurement precision without requiring direct complex hardware modifications.
2Reliability
If obstacle detection is performed without accurate range information, then the system operation is simple, but the reliability of safe driving assistance is reduced
Solution Approach 1:
The patent performs depth estimation and 3D point cloud generation before obstacle detection and classification. By pre-processing the visual data to include accurate depth and spatial information, the system ensures reliable obstacle range measurement is available when making safety decisions, improving reliability through advance preparation of critical data.
Solution Approach 2:
The transformation to 3D point cloud maps provides comprehensive spatial information including distance, volume, and position of obstacles. This dimensional enhancement enables more reliable judgment of obstacle ranges and better-informed safety assistance decisions, directly improving system reliability.
3Measurement precision
If only 2D image processing is used, then the processing speed is fast, but the measurement precision of obstacle characteristics is insufficient
Solution Approach 1:
The patent extracts only the essential depth information needed for obstacle characterization from the full image data. By focusing on generating 3D point cloud maps that capture critical spatial characteristics rather than processing all possible image features, the system achieves accurate obstacle measurement with reasonable processing time.
Solution Approach 2:
The patent efficiently transforms 2D images to 3D representations by leveraging depth estimation models that process image data in a computationally efficient manner. This dimensional transformation provides accurate obstacle characteristics including position, size, and shape while maintaining acceptable processing speeds through optimized algorithms.
Data Source
AI summary
An assistance method of safe driving applied in a vehicle-mounted electronic device obtains RGB images of scene in front of a vehicle, processes the RGB images by a trained depth estimation model, obtains depth images and converts the depth images into three-dimensional (3D) point cloud maps, determines 3D regions of interest therein, and obtains position and size information of objects in the 3D regions of interest. When the position information satisfies a first preset condition and/or the size information satisfies a second preset condition, the presence of obstacles in the 3D regions of interest is determined and controls the vehicle to issue an alarm. When the position information does not satisfy the first preset condition and/or the size information does not satisfy the second preset condition, the 3D regions of interest are determined as obstacle-free, and permitting the vehicle to continue driving.