Rancang Bangun Sistem Deteksi Timbal dalam Air Menggunakan Algoritma Machine Learning dan Photogrammetry

Assegaf, Ali Hasyimi (2026) Rancang Bangun Sistem Deteksi Timbal dalam Air Menggunakan Algoritma Machine Learning dan Photogrammetry. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5025211131-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (6MB) | Request a copy

Abstract

Pencemaran air oleh logam berat, khususnya timbal (Pb), merupakan masalah lingkungan serius karena bersifat toksik dan persisten serta berdampak jangka panjang terhadap kesehatan manusia dan ekosistem perairan. Pemantauan kualitas air konvensional yang masih mengandalkan pengambilan sampel manual dan analisis laboratorium dinilai kurang efisien dari segi waktu, biaya, dan fleksibilitas lapangan, sehingga diperlukan sistem deteksi yang cepat, akurat, dan aplikatif. Tugas akhir ini mengembangkan sistem deteksi konsentrasi timbal (Pb) dalam air berbasis sensor elektronik, pengolahan citra, dan komputasi awan. Sensor diuji melalui pengujian laboratorium berdasarkan parameter sensitivitas, selektivitas, batas deteksi, stabilitas, dan reprodusibilitas, sementara data sensor diproses menggunakan model regresi Long Short-Term Memory (LSTM) dengan nilai koefisien determinasi (R²) sebesar 99,34%, mean absolute error (MAE) sebesar 2,7317 ppm, dan mean squared error (MSE) sebesar 10,9306 ppm². Selain itu, sistem photogrammetry diterapkan menggunakan metode image stitching berbasis Scale-Invariant Feature Transform (SIFT), Oriented FAST and Rotated BRIEF (ORB), dan Accelerated-KAZE (AKAZE) yang dikombinasikan dengan Brute-Force Matcher (BFMatcher) dan Fast Library for Approximate Nearest Neighbors (FLANN), dengan hasil terbaik diperoleh pada metode SIFT dan AKAZE. Seluruh hasil deteksi disimpan pada basis data berbasis cloud dengan tingkat keberhasilan pengiriman data sebesar 100% dan latensi rata-rata 141,35 ms, sehingga sistem dinilai andal untuk pemantauan kualitas air.
=====================================================================================================================================
Water pollution by heavy metals, particularly lead (Pb), poses a serious environmental problem due to its toxic and persistent nature, causing long-term health risks to humans and aquatic ecosystems. Conventional water quality monitoring still relies on manual sampling and laboratory analysis, which are time-consuming, costly, and lack flexibility for rapid field assessment; therefore, a fast, accurate, and field-deployable detection system is urgently needed. This final project develops a lead (Pb) concentration detection system in water based on electronic sensors, image processing, and cloud computing. The sensor was designed to respond selectively to lead and evaluated through laboratory experiments in terms of sensitivity, selectivity, detection limit, stability, and measurement reproducibility. Sensor data were then processed using a Long Short-Term Memory (LSTM) regression model to estimate lead concentration, achieving a coefficient of determination (R²) of 99.34%, a mean absolute error (MAE) of 2.7317 ppm, and a mean squared error (MSE) of 10.9306 ppm². In addition, a photogrammetry system was implemented for environmental image acquisition using image stitching with Scale-Invariant Feature Transform (SIFT), Oriented FAST and Rotated BRIEF (ORB), and Accelerated-KAZE (AKAZE) feature extraction methods combined with Brute Force Matcher (BFMatcher) and Fast Library for Approximate Nearest Neighbors (FLANN), where SIFT and AKAZE showed superior visual quality. All detection results were stored in a cloud-based database with a 100% success rate and an average latency of 141.35 milliseconds, demonstrating reliable performance for water quality monitoring.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	cloud database, kualitas air, LSTM, photogrammetry, sensor elektronik, timbal, cloud database, electronic sensor, lead, LSTM, photogrammetry, water quality.
Subjects:	Q Science > QA Mathematics > QA336 Artificial Intelligence Q Science > QA Mathematics > QA76.87 Neural networks (Computer Science) T Technology > T Technology (General) T Technology > TR Photography > TR810 Aerial photography
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Informatics Engineering > 55201-(S1) Undergraduate Thesis
Depositing User:	Ali Hasyimi Assegaf
Date Deposited:	31 Jan 2026 04:37
Last Modified:	31 Jan 2026 04:37
URI:	http://repository.its.ac.id/id/eprint/131357

Actions (login required)

View Item