Estimasi Ukuran Tubuh Sapi Brahman di BPTU-HPT Sembawa Sumatera Selatan Menggunakan Citra Digital

Abstract

Pengukuran konvensional ukuran tubuh sapi berisiko menimbulkan stres pada ternak dan membahayakan operator, sehingga metode citra digital dikaji sebagai alternatif yang lebih efisien dan noninvasif. Penelitian ini menganalisis akurasi pengukuran ukuran tubuh sapi Brahman berbasis citra digital dibandingkan metode konvensional pada 66 ekor sapi (34 jantan, 32 betina) berumur 12 – 65 bulan di BPTU-HPT Sembawa, Sumatera Selatan. Citra dianalisis menggunakan ImageJ, dengan oeubah panjang badan (PB), tinggi pundak (TP), tinggi pinggul (TPg), lingkar dada (LD), lebar pinggul (LPg), dan dalam dada (DD). Data dianalisis menggunakan uji t berpasangan, korelasi Pearson, dan MAPE. Peubah PB, TP, TPg, dan DD tidak berbeda nyata antarmetode (p=0,05; MAPE 0,90–5,43%; r=0,349–0,989), sedangkan LPg dan LD berbeda sangat nyata (p<0,01; MAPE 17,21–18,32% dan 11,40–13,13%). Akurasi pengukuran digital konsisten pada kedua kelompok umur. Peubah TP memiliki korelasi tertinggi terhadap bobot badan dan menjadi penduga terbaik, meskipun akurasi pendugaan bobot badan masih rendah.

Conventional body measurement of cattle poses risks of stress to the animals and endangers operator safety, thus digital image analysis was examined as a more efficient and non-invasive alternative. This study analyzed the accuracy of digital image-based body measurement in Brahman cattle compared to the conventional method, using 66 cattle (34 males, 32 females) aged 12–65 months at BPTU-HPT Sembawa, South Sumatra. Images were analyzed using ImageJ, with variables of body length (BL), withers height (WH), hip height (HH), chest girth (CG), hip width (HW), and chest depth (CD). Data were analyzed using paired t-test, Pearson correlation, and MAPE. BL, WH, HH, and CD did not differ significantly between methods (p=0.05; MAPE 0.90–5.43%; r=0.349–0.989), whereas HW and CG differed highly significantly (p<0.01; MAPE 17.21–18.32% and 11.40–13.13%). Measurement accuracy was consistent across both age groups. WH showed the highest correlation with body weight and served as the best predictor ,although body weight estimation accuracy remained low