Using Image Analysis Technique for Predicting Light Lamb Carcass Composition

Area of Science:

• Animal Science
• Agricultural Engineering
• Food Science

Background:

• Assessing sheep carcass composition is crucial for quality evaluation.
• Traditional methods can be labor-intensive and subjective.
• Objective, non-invasive techniques are needed for light lamb carcasses.

Purpose of the Study:

• To evaluate the effectiveness of video image analysis (VIA) for determining the tissue composition of light lamb carcasses.
• To develop predictive models for muscle, fat, and bone content using VIA.
• To establish VIA as an objective tool for lamb carcass assessment.

Main Methods:

• Collected photographic images (lateral and dorsal views) of 55 light lamb carcasses.
• Extracted 21 lateral and 29 dorsal measurements (lengths, widths, angles, areas, perimeters).
• Utilized K-Folds stepwise multiple regression to build predictive models for tissue weights and percentages.

Main Results:

• VIA models, using cold carcass weight (CCW) and image data, explained significant variation in tissue weights (K-fold-R²: 0.83–0.98).
• Models for tissue percentages showed slightly lower explained variation (K-fold-R²: 0.41–0.78) but still relied on VIA measurements.
• Prediction accuracy was highest when combining CCW with both lateral and dorsal view measurements.

Conclusions:

• Video image analysis is a highly effective and objective tool for assessing light lamb carcass composition.
• VIA provides valuable data for predicting muscle, fat, and bone content in lamb carcasses.
• This technique offers a non-invasive and efficient alternative for quality assessment in the meat industry.