Iron Scrap Image Classification Using Segmentation to Reduce Labeling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image analysis methods for iron scrap classification face challenges due to the irregular nature of iron scraps, requiring extensive image collection and labeling, and struggle with inconsistent features such as various shapes, textures, and colors, leading to inefficiencies in segmentation and classification.
Innovation Solution
A method utilizing an AI model for image analysis that minimizes the number of images required for collection by employing a segmentation model to extract target iron scraps and a classification model to determine item and grade information, with an evaluation process to enhance accuracy and efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If instance segmentation is used to classify irregular iron scrap, then classification accuracy is improved, but the amount of image collection and labeling required increases significantly
Solution Approach 1:
The patent applies segmentation by dividing the iron scrap classification problem into distinct regions of interest. The system segments the image to identify and focus on specific scrap regions, extracting relevant features while ignoring irrelevant background areas. This segmentation approach enables accurate classification of irregular scrap shapes without requiring exhaustive image collections, as the model learns to focus on critical segmented regions rather than processing entire images.
2Reliability
If extensive image collection and labeling is performed to capture various iron scrap features, then classification performance is improved, but time and resource consumption increases
Solution Approach 1:
The patent implements preliminary action through pre-processing steps that prepare images before main classification. The system performs preliminary segmentation, feature extraction, and region identification on a subset of images to create optimized training data. This preliminary processing reduces the overall data preparation time by pre-identifying key features and regions that will be used during classification, eliminating the need for extensive manual labeling of entire images.
3Ease of manufacture
If the optical system is installed at unloading locations without hardware engineering, then installation ease is improved, but image quality for AI performance deteriorates
Solution Approach 1:
The patent applies self-service by implementing an automated system that captures images directly at the unloading location without requiring complex hardware engineering or manual optical system configuration. The system uses readily available imaging devices and automatically performs segmentation, feature extraction, and classification. This self-service approach maintains installation simplicity while achieving high image quality through software-based optimization rather than complex hardware engineering.
Data Source
AI summary
Disclosed are a method of providing iron scrap classification information through image analysis, and an apparatus for classifying iron scrap and recording medium that performs the method. The method includes obtaining, by a receiving unit, a loaded state image captured in a state in which a plurality of iron scraps are loaded onto a loading device, obtaining, by a processor, segmented images including target iron scrap from the loaded state image using a segmentation model which performs segmentation on the target iron scrap, which is any one of the plurality of iron scraps, obtaining, by the processor, item information and grade information corresponding to the target iron scrap using a classification model which performs classification on the segmented images and performs analysis on the classified images in units of images, and providing, by the processor, iron scrap classification information including the item information and the grade information.