Particle and salinity sensing for the marine environment via deep learning using a Raspberry Pi

Abstract

The identification of mixtures of particles in a solution via analysis of scattered light can be a complex task, due to the multiple scattering effects between different sizes and types of particles. Deep learning offers the capability for solving complex problems without the need for a physical understanding of the underlying system, and hence offers an elegant solution. Here, we demonstrate the application of convolutional neural networks for the identification of the concentration of microparticles (silicon dioxide and melamine resin) and the solution salinity, directly from the scattered light. The measurements were carried out in real-time using a Raspberry Pi, light source, camera, and neural network computation, hence demonstrating a portable and low-cost environmental marine sensor.

Critical Review of Processing and Classification Techniques for Images and Spectra in Microplastic Research

“…164 A tool for automated detection of microplastic particles was later developed using these principles. 165 Because of the ability of CNNs to automatically extract task-relevant information from even complex data sets, deep learning networks can achieve performance similar or superior to humans on specialized tasks. 166,167 However, in order to train deep neural networks large-scale data sets with sufficient manually annotated ground truth data is needed, which remains both time-consuming and costly to obtain.…”

Arı hastalıklarının hibrit bir derin öğrenme yöntemi ile tespiti

“…Fakat derin öğrenme modellerinde ise uzman bir kişiye gereksinim duyulmadan özniteliklerin otomatik olarak çıkartılabilmesi, derin öğrenmeyi popüler hale getirmiştir. Bu nedenle derin öğrenme modelleri ekonomiden eğitime kadar birçok farklı alanda kullanılabilmektedir [47][48][49]. Klasik bir derin öğrenme algoritmasında veriler Şekil 4'de gösterildiği sırayla; giriş, evrişimsel, aktivasyon, havuzlama, ezberleme, tam bağlantı ve sınıflandırma katmanlarından geçirilerek sonuç üretilmektedir.…”

Digital video recorder for Raspberry PI cameras with multi-camera synchronous acquisition

“…One critical criterion is the camera cost. Low cost is one reason the RPi camera has been widely used, especially in prototyping applications [6] , [7] , [8] . Along with low-cost, other factors accounting for the popularity of the RPi in prototyping applications include: (1) RPi has a wide user-base for hobbyists and prototyping applications with excellent on-line support forums, (2) The base RPi board has compute power that can be harnessed for custom video analysis code, (3) The base board has general purpose I/O that are often needed to interface with external signals and systems, and (4) The remote-head camera board can be used with an array of ribbon cable lengths allowing for integration into various mechanical setups.…”