AGV Guidance Using Camera-Based Centroid Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current automated guided vehicles face challenges in accurately following paths marked by colored tape in high traffic areas due to damage, dirt, and color changes, and existing solutions like laser systems are costly and increase the weight and complexity of the vehicles.
Innovation Solution
A guidance system that uses a sensor system to generate images of the environment and identify centroids based on predefined standards, allowing the vehicle to generate commands for navigation, thereby enabling accurate path following without increasing weight, size, or cost.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If laser systems and three-dimensional imaging systems are used to follow predefined paths, then path following accuracy is improved, but cost, weight, and device complexity increase
Solution Approach 1:
The patent replaces complex mechanical/laser imaging systems with a simpler camera-based vision system. The camera captures images of colored tape markers, and image processing algorithms extract path information, substituting expensive laser systems with more affordable and lighter optical components.
Solution Approach 2:
The system creates a visual copy of the path through colored tape markers that can be detected by the camera. Instead of using complex physical guidance systems, the patent uses visual representations (colored tape) that can be easily captured and processed by standard imaging devices.
2Ease of manufacture
If colored tape is used to mark paths, then cost and ease of installation are improved, but reliability deteriorates in high traffic areas due to damage, dirt, and color changes
Solution Approach 1:
The system continuously captures images of the colored tape and processes them to determine the vehicle's position and orientation. This real-time feedback allows the vehicle to adapt to changes in the tape appearance (dirt, damage, color variations) by constantly updating its position based on the latest image data.
Solution Approach 2:
The patent explicitly accounts for color changes in the tape due to environmental factors. The image processing system is designed to recognize the colored tape even when its appearance changes, using color space transformations and pattern recognition to maintain reliable path detection despite dirt, fading, or damage.
3Measurement precision
If image processing with pixel analysis is used, then path following accuracy is maintained, but computational requirements increase
Solution Approach 1:
The system extracts only the essential information from images - specifically, identifying colored tape pixels and calculating their centroid. Rather than processing entire images or using complex three-dimensional reconstruction, the patent extracts minimal necessary data (pixel coordinates and centroid position) to achieve accurate path following with reduced computational load.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method, apparatus and system for guiding a number of mobile devices (106). An image of a portion of an environment (104) around a mobile device (108)in the number of mobile devices (106) is received from a sensor system (114) associated with the mobile device (108). A set of pixels (148) in the image (130) that meets a standard (142) is identified. A centroid (150) for the set of pixels (148) that meets the standard (142) is identified. A number of commands (152) for guiding the mobile device (108) within the environment (104) is generated based on the centroid (150).