In the heart of South Africa’s thriving mohair industry, a groundbreaking study is set to revolutionize how we assess fibre quality. Published in the *South African Journal of Animal Science* (translated from Afrikaans as *South African Journal of Animal Science*), the research, led by N Giovannini from the National Agricultural Technology Institute, compares human expertise with artificial intelligence (AI) in detecting medullated fibres in mohair—a critical factor for the textile industry.
Mohair, a luxurious fibre produced by Angora goats, is a cornerstone of South Africa’s agricultural sector. However, the presence of medullated fibres, which have a distinct core structure, can pose challenges in textile processing. Traditionally, human experts have manually assessed these fibres, a time-consuming and subjective process. Giovannini’s study introduces a more efficient and potentially more accurate method using AI.
The research involved three experiments, each comparing human image analysis to digital image analysis and AI models. The results were striking. “Laboratory personnel showed better coincidence and higher correlations with the AI models when counting objectionable medullated (SME) fibres than when counting industry non-objectionable medullated (NOB) fibres,” Giovannini explained. This discrepancy could be due to the more clearly defined characteristics of SME fibres.
The implications for the mohair industry are profound. AI technologies offer a pathway to modernising fibre quality analysis, enhancing accuracy and efficiency. This aligns with South Africa’s strategic focus on smart agriculture and innovation, potentially boosting the country’s global competitiveness in the mohair market.
Giovannini’s research highlights the growing potential of AI in the agricultural sector. “The results of this study indicate a great advance in the automatic detection of SME and NOB fibres in mohair samples,” Giovannini noted. However, further adjustments to the AI models are required to improve the counting of NOB fibres.
As the world moves towards smarter, more efficient agricultural practices, this research could shape future developments in fibre quality assessment. By integrating AI into the mohair industry, South Africa can lead the way in innovation, ensuring the continued success of this valuable sector.
In the words of Giovannini, “This study opens up new possibilities for the mohair industry, paving the way for more accurate and efficient fibre quality analysis.” As we look to the future, the collaboration between human expertise and AI technologies promises to drive significant advancements in the field.