Double yolk duck egg feature discrimination and size grading based on machine vision and CH-GO rule

• Published in: Journal of food measurement & characterization Vol. 19; no. 3; pp. 1662 - 1672
• Main Authors: Le, Chu Jia, Dan, Liang, Cheng, Wang Jian, Jie, Ye Min
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.03.2025; Springer Nature B.V
• Subjects: Algorithms; Aquatic birds; automatic detection; Chemistry; Chemistry and Materials Science; Chemistry/Food Science; color; computer vision; Convexity; duck eggs; ducks; Eggs; Engineering; Food Science; Image contrast; Image processing; Image segmentation; Machine vision; Original Paper; Segmentation; Vision systems; Yolk; Food detection; Machine vision; Double yolk duck egg; Size grading
• ISSN: 2193-4126; 2193-4134
• DOI: 10.1007/s11694-024-03062-z

The internal irregular yolk shape makes it difficult to identify and grading the double yolk duck eggs accurately. In this paper, a machine vision detection method is proposed to realize the automatic detection of double yolk eggs with different sizes. Firstly, a yolk region extraction algorithm based on green component and threshold segmentation is designed. The yolk image contrast is obtained by extracting single-channel color component, and the preliminary double yolk region is got by adaptive threshold segmentation and morphological processing. Then, the yolk region is divided into separate or connected types, and yolk discrimination rules based on Convex Hull and Green’s Operator (CH-GO) are proposed to discriminate the double yolk features. The dimension parameters of the eggs are extracted and then get graded according to the yolk type and size criterion. Finally, experimental results show that the discrimination rate of double yolk reaches 97.6%, and the size grading accuracy is about 98.4%. The proposed method shows great application potential in the automatic detection and grading of special poultry eggs.