Pig’Scan is an automated image acquisition and analysis system developed for pig slaughter lines.
It combines RFID/tinet identification, synchronized industrial cameras, and image analysis algorithms to characterize carcasses and improve tracking. Three color cameras, protected in waterproof housings and positioned on the line, capture images of the back, head, and genital area of each carcass. The images are associated with a slaughter number to ensure complete traceability.
The system measures tail length, detects tail lesions and bites, and performs automatic sexing of the animals. Data is processed via dedicated software modules for acquisition, visual analysis, and statistical export. The results are structured into usable datasets, including annotated images and CSV files. The goal of Pig’Scan is to standardize carcass evaluation by reducing variability between operators through reproducible digital measurements.
Several methods for estimating tail length have been studied, such as bounding box, base-to-tip measurement, and segmentation approaches based on curvature. These methods improve the distinction between short and long tails under industrial conditions. The modular architecture allows for the integration of new health indicators and the identification of skin marks related to rearing or transport conditions.
Innov'Space 2026 Label !
More infos :
https://www.pigscan.fr or
https://www.unissia.com
