Dump Body Position Specification Using Neural Network Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
At construction sites, the presence of ruts on the ground surface can interfere with the accurate specification of a dump truck's dump body position in images, leading to potential misidentification due to edge detection issues.
Innovation Solution
An image processing system that includes a data acquisition unit, an area-specifying unit, and a loading/unloading target-specifying unit, utilizing a neural network-based segmentation model to accurately identify and specify the dump body within captured images, even in the presence of ruts, by generating three-dimensional data and fitting a target model to improve robustness.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If edge extraction is used to specify dump body position, then the position can be identified from captured images, but ruts on the ground surface generate false edges that cause misidentification
Solution Approach 1:
The patent segments the image processing into multiple stages: first extracting edges, then identifying candidate regions, and finally verifying positions using multiple features (edge continuity, region shape, position relationships). This multi-stage segmentation approach separates the dump body detection from ground surface features, preventing rut interference.
Solution Approach 2:
The patent applies different processing strategies to different regions of the image. Candidate region extraction focuses on specific areas where the dump body is likely to appear, and verification uses local geometric properties and relationships. This localized approach distinguishes the dump body from ruts based on their different local characteristics.
2Reliability
If traditional edge extraction methods are used, then processing is simple and fast, but robustness against ground surface variations is poor
Solution Approach 1:
The processing system is segmented into distinct functional modules: edge extraction unit, candidate region extraction unit, and position specification unit. Each module performs a specific function, making the overall system more reliable while maintaining manageable complexity through modular design.
Solution Approach 2:
The patent introduces candidate region extraction as an intermediary step between edge extraction and final position specification. This intermediate layer filters out false edges from ruts before they reach the position specification stage, improving robustness without requiring complete redesign of the processing pipeline.
Data Source
AI summary
In the image processing system according to the present invention, the data acquisition unit acquires a captured image showing a loading/unloading target of a transporting material of a work machine. The area-specifying unit specifies an area including the loading/unloading target from the captured image. The loading/unloading target-specifying unit specifies at least one predetermined surface of the loading/unloading target from the area including the loading/unloading target.