Khoiriyah, Ummul (2021) Sistem Monitoring Kesehatan Udang Vannamei Menggunakan Image Processing Berbasis Internet of Things (IoT). Undergraduate thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
10511710000012_Undergraduate_Thesis.pdf - Accepted Version Download (2MB) | Preview |
Abstract
Berbagai virus diketahui telah banyak menyerang udang vaname, udang yang terinfeksi virus menunjukkan beberapa hal yang tidak normal, pada Udang Vannamei yang sehat menunjukkan tubuh udang memiliki warna keabuan, dan tidak terdapat warna kemerahan pada ekor dan tubuh udang. Implementasi pembuatan prototipe sistem monitoring kesehatan udang perlu dilakukan untuk mengetahui kondisi kesehatan pada udang. Menyajikan prototipe yang terdiri dari perangkat keras dan perangkat lunak menggunakan metode deep learning YOLO (You Look Only Once) dengan memberikan variasi jarak 20cm dan 30cm mendapatkan akurasi sebesar 81% dan 100% untuk pendeteksian udang sehat, dan variasi 20cm dan 30cm mendapatkan akurasi sebesar 71% dan 92% untuk pendeteksian udang sakit. Pada Histogram diketahui ekstraksi dari udang sehat memiliki rentang pixel 100 sampai 200 mengambarkan histogram dari warna nampan putih tersebut dan pada pixel 50 sampai 80 terdapat kenaikan grafik yang mengambarkan objek udang. Pada udang sakit mendapatkan nilai rentang pixel 120 sampai 200 memiliki pixel tertinggi yang mengambarkan histogram dari warna nampan putih tersebut. Pada pixel 0 grafik naik hingga mencapai 4000 no pixel pada pixel 0 sampai 40 terdapat kenaikan grafik yang menurun hingga pixel 100 memiliki grafik terendah yang mengambarkan objek udang. Hasil Monitoring dapat ditampilkan pada Web Server.
Kata Kunci: Sitem monitoring, Image Processing, Internet of Things (IoT).
=================================================================================================
Various viruses are known to have attacked vaname shrimp, shrimp infected with the virus showed some abnormalities, healthy Vannamei Shrimp shows that the body of the shrimp has a brownish color, the tail fin (Uropoda) expands like a fan, and there is no reddish color on the tail and body of the shrimp. Implementation a prototype consisting of hardware and software using the YOLO (You Look Only Once) deep learning method by providing a distance variation of 20cm and 30cm getting an accuracy of 81% and 100% for the detection of healthy shrimp, and variations of 20cm and 30cm getting an accuracy of 71 % and 92% for the detection of sick shrimp. In the Histogram, it is known that the extraction from healthy shrimp has a pixel range of 100 to 200 which represents the histogram of the color of the white tray and at pixels 50 to 80 there is an increase in the graph depicting the shrimp object and in sick shrimp, the pixel value ranges from 120 to 200 which has the highest pixel which represents the histogram. of the color of the white tray. At pixel 0 the graph rises to 4000 no pixels, at pixels 0 to 40 there is an increase in the graph that decreases until pixel 100 has the lowest graph depicting a shrimp object. Monitoring can be displayed on the Web Server.
Keywords: Monitoring System, Image Processing, Internet of Things (IoT).
| Item Type: | Thesis (Undergraduate) |
|---|---|
| Uncontrolled Keywords: | Sitem Monitoring, Image Processing, Internet of Things (IoT),Monitoring System, Image Processing, Internet of Things (IoT) |
| Subjects: | Q Science > QA Mathematics > QA336 Artificial Intelligence T Technology > TA Engineering (General). Civil engineering (General) > TA1637 Image processing--Digital techniques. Image analysis--Data processing. |
| Divisions: | Faculty of Vocational > Instrumentation Engineering |
| Depositing User: | Ummul Khoiriyah |
| Date Deposited: | 24 Aug 2021 07:01 |
| Last Modified: | 14 Nov 2024 05:21 |
| URI: | http://repository.its.ac.id/id/eprint/90053 |
Actions (login required)
 |
View Item |