RGB-D Carpet Detection for Mobile Robot Curl Avoidance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing carpet detection methods for mobile robots are inefficient and costly, particularly in detecting low-height carpets and carpet-curls, which can lead to dangerous situations, and do not adequately optimize navigation in scenarios with carpets.
Innovation Solution
A carpet detection method using an RGB-D camera with deep learning models to detect both carpets and carpet-curls, allowing for effective navigation by generating bounding boxes and points to identify and avoid obstacles, integrated with a navigation module for trajectory planning and collision avoidance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a combination of many sensors is used to detect carpets, then the detection capability is improved, but the system complexity and cost increase
Solution Approach 1:
The patent applies multi-functionality by enabling a single RGB-D camera to perform multiple detection tasks: detecting both flat carpets and carpet-curls, while also providing depth information and visual data for navigation. This replaces the need for multiple specialized sensors, resolving the contradiction between detection capability and system complexity
Solution Approach 2:
The patent changes the detection parameters by using an RGB-D camera that captures both color (RGB) and depth (D) information simultaneously. This parameter expansion allows the single sensor to detect carpets through multiple characteristics (color patterns and depth variations), improving detection capability without adding more sensors
2Device complexity
If existing sensor methods are used for carpet detection, then the system structure is simple, but the detection efficiency and accuracy for low-height carpets and carpet-curls deteriorate
Solution Approach 1:
The patent transitions from 2D visual detection to 3D spatial detection by incorporating depth information from the RGB-D camera. This dimensional enhancement enables the detection of low-height carpets and carpet-curls that are invisible to conventional 2D sensors, improving detection efficiency while maintaining relatively simple system structure
Solution Approach 2:
The patent replaces mechanical contact-based carpet detection (such as bump sensors) with optical-depth sensing using an RGB-D camera. This substitution enables non-contact, real-time detection of carpets and carpet-curls, significantly improving detection efficiency and responsiveness while maintaining simple system structure
3Ease of operation
If existing carpet detection methods are used, then the implementation is straightforward, but the navigation safety in scenarios with carpets and carpet-curls deteriorates
Solution Approach 1:
The patent implements preliminary action by detecting and mapping carpets and carpet-curls in advance during the exploration and mapping phase. This allows the navigation system to pre-plan trajectories that avoid detected carpet-curls and adjust speed when approaching carpets, improving navigation safety before the robot encounters these obstacles
Solution Approach 2:
The patent implements feedback mechanisms where the detected carpet and carpet-curl information is continuously fed back to the trajectory planning and speed control modules. This real-time feedback enables dynamic adjustment of navigation parameters, improving navigation safety while maintaining straightforward implementation through existing control architectures
Data Source
AI summary
Carpet detection using an RGB-D camera and mobile machine movement control based thereon are disclosed. Carpets and carpet curls are detected by obtaining a RGB-D image pair including an RGB image and a depth image through an RGB-D camera, detecting carpet and carpet-curl areas in the RGB image and generating a 2D bounding box to mark each area using a deep learning model, and generating groups of carpet and carpet-curt points corresponding to each of the carpet and carpet-curl areas by matching each pixel of the RGB image within each 2D bounding box corresponding to the carpet and carpet curl areas to each pixel in the depth image.