Tuber Ruler: a mobile application for evaluating image-based potato tuber size

In the potato processing industry, the length-to-width (L/W) ratio of potato tubers is a critical quality indicator, especially for products like French fries and chips. Traditional measurement methods such as manual scales or calipers are labor-intensive and subject to variability. Addressing this challenge, we introduce a mobile application (Tuber Ruler) developed for Android smartphones, which employs image analysis to accurately measure the L/W ratio of potato tubers. By integrating standard image-processing and machine learning (ML) techniques, Tuber Ruler offers a dual-approach analysis, allowing for rapid and precise tuber sizing against both standard black and diverse natural backgrounds. The application exhibits acceptable performance metrics when compared to ground truth measurements obtained via digital calipers. Specifically, Tuber Ruler achieved a Pearson’s correlation coefficient ( r ) of 0.98 and a mean absolute error (MAE) of 0.03 for russet potatoes, demonstrating consistent accuracy across diverse potato varieties (yellow tubers: r = 0.98, MAE = 0.02; red tubers: r = 0.91, MAE = 0.04). The app can be used in challenging natural environments such as soil and grass backgrounds. Moreover, Tuber Ruler maintains high accuracy ( r ≥ 0.99, MAE = 0.01–0.04) even when image resolution is reduced to 25% of the original size, showcasing its resilience to resolution degradation. A significant aspect of Tuber Ruler is the effective use of ML to complement standard image-processing approach, enhancing the application’s adaptability to varied backgrounds and tuber types. This dual-approach analysis, coupled with swift processing times (standard image-processing: 2.0 s; ML: 4.0 s), positions Tuber Ruler as an alternative to traditional sizing methods. By offering a scalable, precise, and user-friendly tool for tuber sizing, Tuber Ruler has the potential to significantly enhance productivity and operational efficiency in the potato industry, becoming a valuable tool to growers, processors and researchers.