Curl type determination and pelt size estimation of new born lambs of Zandi sheep Using Artificial Intelligence techniques

In this study, a method based on using image processing and artificial neural network is introduced to determine curl type and pelt size of new born lambs in Zandi sheep. The data was collected from luxury pelt collection belong to the Zandi sheep breeding center Khojir, Tehran, Iran. Primarily, curl type and dimensions were determined by experienced appraisers for some pelts. Then, some digital images were captured from different curl types on the pelts and from different pelt dimensions. The features related to curl type determination and pelt size estimation were extracted from digital images using image processing tools (IPT) of MATLAB software. Two different Artificial Neural Networks were designed for classification of curl types and estimation of pelt size, independently. To classifying the curl types, a neural network was trained with the accuracy of 100%, and in the test phase it could classify the 7 curl types with the accuracy of near 95%. The accuracy of pelt size estimation was near 99% for the second tested neural network. There was a significant correlation (97.15%) between true and ANN estimated pelt sizes. The results indicated that there is a potential of using artificial intelligence as a replacement for the human sense, to determine pelt quality in the new born lambs of pelt sheep.