Shelf Edge Detection Using Hough Transform and Depth Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In complex environments like retail and warehousing facilities, detecting structural features such as shelf edges is challenging due to varying object positions, orientations, lighting conditions, and label placements, which hinders accurate navigation and inventory management.
Innovation Solution
A mobile automation system equipped with image sensors and depth sensors uses a Hough transform to detect preliminary edges, segments depth measurements to identify seed pixels, and applies a line-fitting model to estimate shelf edges, enhancing edge detection accuracy and robustness against calibration errors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional edge detection methods are used in complex retail environments, then the system can operate with simple detection algorithms, but the detection accuracy deteriorates due to varying object positions, orientations, lighting conditions, and label placements
Solution Approach 1:
The patent segments the detection process into multiple independent modules: color space conversion module, edge detection module, Hough transform module, and line-fitting module. Each module handles a specific aspect of the detection task, allowing the system to achieve high accuracy through coordinated processing of multiple features rather than relying on a single complex algorithm
Solution Approach 2:
The patent transforms the detection problem from 2D image space to 3D space by integrating depth information from depth sensors. The line-fitting model operates in three-dimensional space using coordinates (x, y, z) to represent shelf edges, adding a depth dimension that provides additional geometric constraints and improves detection accuracy in complex environments
2Measurement precision
If the system uses multiple sensors and processing steps to improve detection accuracy, then measurement precision improves, but the processing time and computational load increase
Solution Approach 1:
The patent performs preliminary color space conversions and edge detections before applying the Hough transform. By pre-processing the image data to identify potential edge regions and convert to appropriate color spaces, the system reduces the computational burden of subsequent processing steps and enables faster execution of the more computationally intensive line-fitting operations
Solution Approach 2:
The patent replaces traditional mechanical edge detection approaches with computational methods including Hough transform and line-fitting models. This substitution allows the system to achieve high precision through mathematical transformations and statistical fitting rather than relying on simple threshold-based or gradient-based detection, balancing accuracy with processing efficiency
3Measurement precision
If the system applies calibration requirements to ensure accuracy, then measurement precision improves, but the ease of operation deteriorates due to calibration errors and setup complexity
Solution Approach 1:
The patent changes the parameter space by working with normalized coordinates and transformed color spaces rather than raw sensor data. The line-fitting model uses normalized three-dimensional coordinates and transforms edge detection results into a standardized parameter space, making the system less sensitive to calibration errors and reducing the impact of varying lighting conditions and sensor misalignments
Data Source
AI summary
A method of detecting an edge of a support surface in an imaging controller includes: obtaining image data captured by an image sensor and a plurality of depth measurements captured by a depth sensor, the image data and the plurality of depth measurements corresponding to an area containing the support surface; detecting preliminary edges in the image data; applying a Hough transform to the preliminary edges to determine Hough lines representing candidate edges of the support surface; segmenting the plurality of depth measurements to assign classes to each pixel, each class defined by one of a plurality of seed pixels, wherein the plurality of seed pixels are identified from the depth measurements based on the Hough lines; and detecting the edge of the support surface by selecting a class of pixels and applying a line-fitting model to the selected class to obtain an estimated edge of the support surface.