研究目的
To develop a machine vision system for the automated grading of opals, addressing the subjectivity and inconsistency in current opal evaluation practices.
研究成果
The GDA system, based on imaging and statistical learning, provides a more consistent and objective method for opal grading compared to manual evaluations. The system has potential to be extended to grade other gemstones.
研究不足
The limited number of opals available for the experiments may compromise the results. More opal samples from various classes and grades within each class would help improve experimental results.
1：Experimental Design and Method Selection:
The GDA system is designed to mimic human observation of an opal by rotating it under controlled lighting and capturing its images. The system uses statistical machine learning for opal classification and grading.
2：Sample Selection and Data Sources:
Opals are scanned over the full range of viewing angles with different lighting angles and camera exposures, capturing 865 images per opal.
3：List of Experimental Equipment and Materials:
The GDA hardware includes a two-axis translucent stage, white LED lighting, a 5-megapixel camera, overhead and angular halogen lights, and a suction system to secure the opal.
4：Experimental Procedures and Operational Workflow:
The workflow includes calibration, opal image capture, image analysis, and opal classification and grading.
5：Data Analysis Methods:
Image analysis techniques are used to extract opal features, and statistical machine learning models are employed for classification and grading.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
