Virtual Horizontal Stereo Camera for Vertical Object Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Stereo vision systems face challenges in accurately detecting vertical objects due to singularity issues in vertical baseline orientations, and implementing multiple orientations is cost-prohibitive, limiting effective distance calculation and spatial relationship determination.
Innovation Solution
A virtual horizontal stereo camera system is developed, comprising a vertically oriented stereo camera pair and a processor that generates a virtual horizontal disparity image using convolutional neural networks, enabling object detection and improving the detection of thin vertical objects without the need for a physically horizontal stereo camera.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If a vertical baseline orientation stereo camera is used, then the system can be implemented with a single stereo pair, but it fails to accurately detect vertical objects due to singularity in disparity calculation
Solution Approach 1:
The patent creates a virtual horizontal disparity image that copies the function of a physical horizontal stereo camera. By using image processing and computational algorithms, the system generates synthetic disparity information that mimics what a horizontal baseline camera would produce, eliminating the need for expensive multiple stereo pairs while maintaining detection accuracy for vertical objects
Solution Approach 2:
The patent introduces an intermediary processing layer that transforms vertical baseline disparity data into horizontal disparity representations. This intermediary virtual horizontal disparity image acts as a mediator between the vertical camera configuration and the requirements for detecting vertical objects, resolving the contradiction without requiring physical hardware changes
2Measurement precision
If a horizontal baseline orientation stereo camera is used, then vertical objects can be detected accurately, but the cost increases due to needing multiple stereo camera pairs
Solution Approach 1:
Instead of physically implementing expensive horizontal stereo camera pairs, the patent creates a virtual copy of horizontal disparity information through computational processing of vertical baseline images. This digital copying approach achieves the same detection capability at fraction of the hardware cost
Solution Approach 2:
The patent replaces the mechanical solution of adding physical horizontal stereo cameras with a computational approach. By using image processing algorithms and neural networks, the system substitutes mechanical hardware expansion with software-based virtual camera generation, significantly reducing costs while maintaining accuracy
3Adaptability or versatility
If multiple stereo camera pairs with different orientations are implemented, then both vertical and horizontal object detection is improved, but the device complexity and cost become prohibitive
Solution Approach 1:
The patent makes a single vertical baseline stereo camera system perform multiple functions by generating both vertical and virtual horizontal disparity images. This multi-functional approach allows the same hardware to achieve detection capabilities previously requiring multiple specialized camera pairs, reducing complexity while maintaining versatility
Solution Approach 2:
The patent adds a computational dimension to the physical camera system. By processing vertical baseline images through algorithms that generate virtual horizontal disparity, the system effectively adds a second orientation dimension without physical cameras, achieving multi-orientation detection capability without increasing hardware complexity
Data Source
AI summary
An apparatus including a stereo camera and a processor. The stereo camera may comprise a first capture device and a second capture device in a vertical orientation. The first capture device may be configured to generate first pixel data and the second capture device may be configured to generate second pixel data. The processor may be configured to receive the first pixel data and the second pixel data, generate a vertical disparity image in response to the first pixel data and the second pixel data, generate a virtual horizontal disparity image in response to the first pixel data and the vertical disparity image and detect objects by analyzing the vertical disparity image and the virtual horizontal disparity image. An analysis of the virtual horizontal disparity image may enable the processor to detect the objects not detected in the vertical disparity image alone.