Please use this identifier to cite or link to this item:
http://repository.ipb.ac.id/handle/123456789/117696| Title: | Pengembangan Alat Penduga Kandungan Klorofil Daun Secara Non-Destruktif Berbasis Multi-Channel Visible Light Sensor |
| Other Titles: | Development of Non-Destructive Leaf Chlorophyll Content Prediction Device Based on Multi-Channel Visible Light Sensor |
| Authors: | Widodo, Slamet Reza, Rizki Muhamad |
| Issue Date: | 15-May-2023 |
| Publisher: | IPB University |
| Abstract: | Pendugaan kandungan klorofil yang tersedia saat ini sangat mahal dan sulit
dijangkau petani menengah ke bawah. Pendugaan klorofil penting untuk
mengetahui kebutuhan pupuk nitrogen yang dibutuhkan tanaman. Penelitian ini
bertujuan mengembangkan alat penduga kandungan klorofil yang murah sebagai
alternatif SPAD-502Plus. Alat yang dikembangkan menggunakan sensor AS7262
dan Arduino UNO sebagai mikrokontroler. Proses kalibrasi alat dilakukan dengan
membandingkan hasil pembacaan sensor AS7262 dengan hasil pengukuran
chlorophyll meter SPAD-502Plus Konica Minolta. Proses kalibrasi dilakukan
menggunakan sampel berupa daun kopi, cabai, dan kakao masing-masing 200
lembar. Hasil kalibrasi menunjukkan adanya korelasi antara nilai SPAD dengan
nilai sensor, terlebih pada channel kuning. Dilakukan analisa dengan dua metode,
single channel dan multi-linear regression. Masing-masing metode menghasilkan
empat persamaan yang kemudian disematkan dalam program Arduino dan
dilakukan pengujian alat. Pengujian dilakukan pada delapan persamaan
menggunakan sampel daun masing-masing 50 lembar daun. Pengujian pada
persamaan hasil analisa multi-linear regression menghasilkan hasil yang terbaik.
Pengujian persamaan kopi menghasilkan R2 = 0,9872; RMSE = 1,79; dan MAPE =
2,90%. Pengujian persamaan cabai menghasilkan R2 = 0,9725; RMSE=1,57; dan
MAPE = 3,76%. Pengujian persamaan kakao menghasilkan R2 = 0,9828; RMSE =
1,40; dan MAPE = 3,32%. Pengujian persamaan untuk tanaman umum
menghasilkan R2 = 0,9518; RMSE = 3,16; dan MAPE = 7,04%. Seluruh pengujian
mengindikasikan pendugaan kandungan klorofil menggunakan alat yang
dikembangkan pada penelitian ini sangat akurat. The present techniques for estimating chlorophyll content come with a high price tag and present formidable obstacles for farmers from middle to lower socioeconomic backgrounds. Chlorophyll estimation is important to determine the nitrogen fertilizer requirements needed by plants. This research aims to develop a cheap chlorophyll content estimator as an alternative to SPAD-502Plus. The developed tool uses AS7262 sensor and Arduino UNO as a microcontroller. The tool calibration process is done by comparing the AS7262 sensor reading results with the measurement results of the SPAD-502Plus Konica Minolta chlorophyll meter. The calibration process was carried out using 200 samples of coffee, chili, and cocoa leaves each. Calibration results show a correlation between SPAD values and sensor values, especially in the yellow channel. The data were analyzed using two methods, single channel and multilinear regression. Each method produces four equation models which are then embedded in the Arduino program and tested. Tests were conducted on eight equations using leaf samples of 50 leaves each. Testing on the equation of multilinear regression analysis results in the best results. Testing the coffee equation produced R2 = 0.9872; RMSE = 1.79; and MAPE = 2.90%. Testing the chili pepper equation resulted in R2 = 0.9725; RMSE = 1.57; and MAPE = 3.76%. Testing the cocoa equation resulted in R2 = 0.9828; RMSE = 1.40; and MAPE = 3.32%. Testing the equation for general crops resulted in R2 = 0.9518; RMSE = 3.16; and MAPE = 7.04%. All tests indicated that the estimation of chlorophyll content using the tool developed in this study was very accurate. |
| URI: | http://repository.ipb.ac.id/handle/123456789/117696 |
| Appears in Collections: | UT - Agricultural and Biosystem Engineering |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| Cover_F14180111_Rizki Muhamad Reza.pdf Restricted Access | Cover | 423.86 kB | Adobe PDF | View/Open |
| F14180111_Rizki Muhamad Reza.pdf Restricted Access | Fulltext | 3.21 MB | Adobe PDF | View/Open |
| Lampiran_F14180111_Rizki Muhamad Reza.pdf Restricted Access | Lampiran | 799.04 kB | Adobe PDF | View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.