Estimasi Bobot Ternak Sapi Dengan Metode Deformable Template Dan Klasifikasi Support Vector Machine Multiclass

Abstract

AbstrakSapi adalah hewan ternak anggota family Bovidaedan sub family Bovinae[1]. Sapi dipelihara terutama untuk dimanfaatkan susu dan dagingnya sebagai pangan manusia. Sapi merupakan komoditas peternakan yang banyak dijual-belikan. Seiring dengan pertumbuhan penduduk yang semakin pesat, permintaan produk dari sapi pun juga meningkat terutama dalam hal permintaan daging, susu, maupun kulit. Hasil dari produk sapi dipengaruhi oleh perawatan sapi dan bobot sapi. Dalam melakukan penimbangan bobot badan ternak sapi masih banyak dilakukan dengan cara konvensional. Apabila setiap kali melakukan penimbangan memakai cara konvensional, tentu kurang praktis. Pada tugas akhir ini telah dibuat aplikasi berbasis Matlab untuk membantu mengetahui bobot ternak sapi dengan menggunakan metode pengolahan citra, yang dilengkapi dengan registrasi citra berbasis metode Deformable Templatedengan klasifikasi Multiclass Support Vector Machine (SVM). Didapatkan tingkat akurasi sebesar 76,1905%. Hasil penelitian ini diharapkan dapat membantu pelaku bisnis ternak sapi dalam standar akurasi yang tepat dalam mengetahui bobot ternak sapi. Kata Kunci: Sapi Penggemukan, Pengolahan Citra,Deformable Template, Multiclass Support Vector Machine (SVM).AbstractCows are livestock members of the Bovidae family and the Bovinae sub family [1]. Cows are raised mainly for their use of milk and meat as human food. Cows are commodities that are widely traded. Along with the increasingly rapid population growth, the demand for products from cows has also increased, especially in terms of demand for meat, milk and skin. The yield of cow products is influenced by cattle care and cattle weight. In carrying out the weighing of cattle body weight, there are still many conventional methods. If every time you weigh using conventional methods, it is certainly not practical.In this final project Matlab-based application has been made to help determine the weight of cattle using image processing methods, which are equipped with registration methods based on the Deformable Template method using the Multiclass Support Vector Machine (SVM) classification.Obtained an accuracy rate of 76,1905%. The results of this study are expected to be able to help cattle business players in the right standard of accuracy in knowing the weight of cattle.Keywords: Fattening Cattle, Image Processing, Image Registration, Deformable Templates, Multiclass Support Vector Machine (SVM).