Mukarromah, Aminatul (2015) Deteksi Siklus Estrus Sapi melalui Analisis Tekstur dan Warna Citra Vulva Sapi menggunakan Learning Vector Quantization. Undergraduate thesis, Universitas Islam Negeri Maulana Malik Ibrahim.
![]() |
Text (Fulltext)
11640019.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Attribution Non-commercial No Derivatives. Download (4MB) | Request a copy |
Abstract
ABSTRAK
Peningkatan produksi susu sapi perah di Indonesia masih rendah jika dibandingkan dengan kebutuhan konsumsi susu secara nasional sehingga Indonesia melakukan kebijakan impor susu dari luar negeri. Oleh karena itu, perlu dilakukan penguatan sistem pembibitan yang benar. 70% Penyebab kegagalan sapi bunting adalah kesalahan deteksi birahi (estrus). Ciri-ciri sapi estrus dapat ditandai dengan warna vulva memerah, menebal, dan hangat. Tujuan dari penelitian ini adalah: (1) Menghasilkan metode baru untuk mendeteksi siklus estrus sapi dengan memanfaatkan citra vulva sapi menggunakan learning vector quantization (LVQ). (2) Mengetahui tingkat akurasi keberhasilan deteksi siklus estrus sapi melalui analisis tekstur dan warna citra vulva sapi dengan LVQ. Tahapan penelitian dibagi 3 yaitu tahapan pengolahan citra, tahapan metode learning vector quantization, dan pembuatan aplikasi deteksi siklus estrus sapi. Tahap pengolahan citra menggunakan analisis warna dan tekstur menghasilkan nilai pixel red, green, blue, mean, entropy, skewness, variance, kurtosis. Data hasil pengolahan citra dan hasil pengukuran suhu vulva sapi digunakan sebagai inputan jaringan LVQ. Proses pembuatan jaringan LVQ dilakukan dengan variasi parameter learning rate untuk mendapatkan hasil yang optimal. Pembuatan aplikasi deteksi siklus estrus sapi untuk menguji tingkat keberhasilan jaringan dalam mengenali siklus estrus sapi. Hasil pengujian akurasi jaringan LVQ dalam mengenali citra vulva sapi estrus dan diestrus sebesar 100%, citra vulva sapi proestrus 60%, dan citra vulva sapi metestrus 0%. Hasil penelitian ini menunjukkan bahwa jaringan LVQ hanya dapat mengenali citra vulva sapi estrus dan diestrus.
ABSTRACT
The increasing of milk production of cattle in Indonesia is still lower when it is compared with the needs of the national milk consumption that makes Indonesia deciding to do the policy of importing milk from abroad. Therefore, it is necessary to strengthen the correct seeding system. Around 70% of the failure cause of pregnant cows is the error of estrus detection. The characteristics of estrus cattle are able to be characterized as follow: the color of vulva is red, thickened, and warm. The aims of this study are: (1) to generate a new method for detecting the estrus cycle of the cattle by utilizing the sketch of the cattle’s vulva using the learning vector quantization; (2) to know the level of the success of the cattle’s estrus cycle detection through the analysis of texture and cattle’s sketch color using LVQ. The stages of this study are divided into 3 stages, such as: sketch processing, the stage of LVQ method, and producing the application detection of the cattle’s estrus cycle. In processing the sketch, the researcher uses the nalysis of color and texture which results pixel RGB values, mean, entrophy, skewness, variance, kurtosis. The data from the sketch processing and the result of the temperature measurement of cattle’s vulva are used as input of LVQ. The process of manufacturing LVQ network is done with a variety of learning rate parameter to obtain the optimal result. The application production of the cattle’s estrus cycle detection is to test the success rate in identifying the cattle’s estrus cycle. The result of accuracy testimonial of LVQ network in recognizing the sketch of the cattle’s estrus cycle and diestrus is 100%, the sketch of cattle’s vulva proestrus is 60%, and the sketch of cattle’s vulva metestrus 0%. The result of this study indicates that the network of LVQ is only able to recognize the sketch of estrus cattle and diestrus cattle.
Item Type: | Thesis (Undergraduate) | |||||||||
---|---|---|---|---|---|---|---|---|---|---|
Supervisor: | Mulyono, Agus and Syarifah, Umaiyatus | |||||||||
Contributors: |
|
|||||||||
Keywords: | Siklus Estrus Sapi; Analisis Tekstur dan Warna; Jaringan Syaraf Tiruan; Learning Vector Quantization Estrus Cycle in Cattle; Texture and Color Image Analysis; Artificial Neuron Network; Learning Vector Quantization | |||||||||
Departement: | Fakultas Sains dan Teknologi > Jurusan Fisika | |||||||||
Depositing User: | Moch. Nanda Indra Lexmana | |||||||||
Date Deposited: | 03 May 2023 13:56 | |||||||||
Last Modified: | 03 May 2023 13:56 | |||||||||
URI: | http://etheses.uin-malang.ac.id/id/eprint/49576 |
Downloads
Downloads per month over past year
Actions (login required)
![]() |
View Item |