Object Detection Device Using Hough Transform on UD Map
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing object detection systems in mobile objects, such as vehicles, face challenges in accurately detecting objects parallel to the direction of travel using stereo cameras, particularly in distinguishing between road surface features and parallel objects like guardrails or vehicles, leading to potential misclassification and reduced detection accuracy.
Innovation Solution
The implementation of a processor-based system that applies a Hough transform to a U-disparity map to detect straight lines and convert them into Hough space, allowing for the identification of target disparities corresponding to objects parallel to the camera's direction of travel, thereby enhancing object detection performance by differentiating between road surface features and parallel objects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a stereo camera is used to detect objects on the road surface, then object detection capability is provided, but accuracy in distinguishing parallel objects from road surface features deteriorates
Solution Approach 1:
The patent segments the disparity detection process into two distinct pathways: one for detecting parallel objects using Hough transform on UD map coordinate points, and another for detecting road surface features using traditional disparity methods. This segmentation allows each pathway to be optimized independently, with the parallel object detection pathway achieving higher precision by focusing computational resources on identifying linear patterns in the disparity space that correspond to parallel objects.
Solution Approach 2:
The patent transforms the two-dimensional UD map into Hough space, adding a dimensional transformation that enables better separation of parallel objects from road surface features. By converting coordinate points from the (u, d) disparity space into Hough space parameters (rho, theta), the system creates a new representation where parallel lines in the image space become points in Hough space, making it easier to distinguish them from the curved patterns representing road surface features.
2Measurement precision
If traditional disparity detection is used, then road surface detection is effective, but parallel objects are misclassified as road surface features
Solution Approach 1:
The patent introduces Hough transform as an intermediary processing step between raw disparity data and final object classification. The Hough transform acts as a mediator that converts disparity coordinate points into a parameter space where parallel objects can be reliably identified through line detection, while road surface features maintain their characteristic curved patterns. This intermediary transformation prevents misclassification by providing a clear distinction between parallel object patterns and road surface patterns.
3Reliability
If Hough transform is applied to UD map, then parallel objects can be distinguished, but processing complexity increases
Solution Approach 1:
The patent performs preliminary filtering and selection of coordinate points from the UD map before applying the Hough transform. By pre-processing the disparity data to identify and select only those coordinate points that are likely to represent parallel objects (based on their spatial distribution and disparity characteristics), the system reduces the computational burden of the subsequent Hough transform while maintaining its effectiveness in distinguishing parallel objects.
Data Source
AI summary
An object detection device includes a processor. The processor is configured to apply a Hough transform to coordinate points on a UD map to detect a straight line having a predetermined length, and detect a target disparity corresponding to the detected straight line having the predetermined length as a disparity corresponding to an object parallel to a direction of travel of a stereo camera. In the UD map, a target disparity satisfying a predetermined condition among disparities acquired from a captured image is associated with coordinate points, each of the coordinate points having two-dimensional coordinates formed by a first direction and a direction corresponding to a magnitude of a disparity. The processor is configured to convert, in the Hough transform, a straight line passing through coordinate points associated with the target disparity and a predetermined range based on a vanishing point into a Hough space.