The natural phenomenon of harmful algae bloom (HAB) has a bad impact on the quality of pure and freshwater. It increases the risk to human health, water bodies and overall aquatic ecosystem. It is necessary to continuously monitor and perform proper action against HAB. The inspection of algae blooms by using conventional methods, like algae detection under microscopes, is a difficult, expensive, and time-consuming task, however, computer vision-based deep learning models play a vital role in identifying and detecting harmful algae growth in aquatic ecosystems and water reservoirs. Many studies have been conducted to address harmful algae growth by using a CNN based model, however, the YOLO model is considered more accurate in identifying the algae. This advanced deep learning method is extensively used to detect algae and classify them according to their corresponding category. In this study, we used various versions of the convolution neural network (CNN) based on the You Only Look Once (YOLO) model. Recently YOLOv5 has been getting more attention due to its performance in real-time object detection. We performed a series of experiments on our custom microscopic images dataset by using YOLOv3, YOLOv4, and YOLOv5 to detect and classify the harmful algae bloom (HAB) of four classes. We used pre-processing techniques to enhance the quantity of data. The mean average precision (mAP) of YOLOv3, YOLOv4, and YOLO v5 is 75.3%, 83.0%, and 91.0% respectively. For the monitoring of algae bloom in freshwater, computer-aided based systems are very helpful and effective. To the best of our knowledge, this work is pioneering in the AI community for applying the YOLO models to detect algae and classify from microscopic images.

/pdf/computer-vision-based-deep-learning-approach-for-the-3pbhf4vg.pdf

Computer Vision Based Deep Learning Approach for the Detection and Classification of Algae Species Using Microscopic Images

Algal blooms have various effects on drinking water supply systems; thus, proper monitoring is essential. Traditional visual identification using a microscope is a time-consuming method and requires extensive labor. Recently, advanced machine learning algorithms have been increasingly applied for the development of object detection models. The You-Only-Look-Once (YOLO) model is a novel machine learning algorithm used for object detection; it has been continuously improved in newer versions, and a tiny version of each standard model presented. The tiny versions applied a less complicated architecture using a smaller number of convolutional layers to enable faster object detection than the standard version. This study compared the applicability of the YOLO models for algal image detection from a practical aspect in terms of classification accuracy and inference time. Therefore, automated algal cell detection models were developed using YOLO v3 and YOLO v4, in which a tiny version of each model was also applied. The cell images of 30 algal genera were used for training and testing the models. The model performances were compared using the mean average precision (mAP). The mAP values of the four models were 40.9, 88.8, 84.4, and 89.8 for YOLO v3, YOLO v3-tiny, YOLO v4, and YOLO v4-tiny, respectively, demonstrating that YOLO v4 is more precise than YOLO v3. The tiny version models presented noticeably higher model accuracy than the standard models, allowing up to ten times faster object detection time. These results demonstrate the practical advantage of tiny version models for the application of object detection with a limited number of object classes.

/pdf/deep-learning-based-algal-detection-model-development-er3ir6me.pdf

Deep Learning-Based Algal Detection Model Development Considering Field Application

Microalgae are essential parts of marine ecology, and they play a key role in species balance. Microalgae also have significant economic value. However, microalgae are too tiny, and there are many different kinds of microalgae in a single drop of seawater. It is challenging to identify microalgae species and monitor microalgae changes. Machine learning techniques have achieved massive success in object recognition and classification, and have attracted a wide range of attention. Many researchers have introduced machine learning algorithms into microalgae applications, and similarly significant effects are gained. The paper summarizes recent advances based on various machine learning algorithms in microalgae applications, such as microalgae classification, bioenergy generation from microalgae, environment purification with microalgae, and microalgae growth monitor. Finally, we prospect development of machine learning algorithms in microalgae treatment in the future.

/pdf/machine-learning-for-microalgae-detection-and-utilization-3q4mkera.pdf

Machine learning for microalgae detection and utilization

Durum wheat milling is a key process step to improve the quality and safety of final products. The aim of this study was to characterize three bran-enriched milling fractions (i.e., F250, G230 and G250), obtained from three durum wheat grain samples, by using an innovative micronization and air-classification technology. Milling fractions were characterized for main standard quality parameters and for alveographic properties, starch composition and content, phenolic acids, antioxidant activity and ATIs. Results showed that yield recovery, ash content and particle size distributions were influenced either by the operating conditions (230 or 250) or by the grain samples. While total starch content was lower in the micronized sample and air-classified fractions, the P/L ratio increased in air-classified fractions as compared to semolina. Six main individual phenolic acids were identified through HPLC-DAD analysis (i.e., ferulic acid, vanillic acid, p-coumaric acid, sinapic acid, syringic and p-hydroxybenzoic acids). Compared to semolina, higher contents of all individual phenolic components were found in all bran-enriched fractions. The highest rise of TPAs occurred in the F250 fraction, which was maintained in the derived pasta. Moreover, bran-enriched fractions showed significant reductions of ATIs content versus semolina. Overall, our data suggest the potential health benefits of F250, G230 and G250 and support their use to make durum-based foods.

Qualitative Characterization of Unrefined Durum Wheat Air-Classified Fractions

Benchmarking analysis of CNN models for pits of sour cherry cultivars

Microalgae are single-celled organisms that have been extensively utilized in biotechnology, pharmacology and foodstuff in recent years. The description and classification of many existing microalgae groups are carried out with classical methods in a long time and with a remarkably qualified labor force. Deep learning methods have achieved success in many fields are applied to the classification of microalga groups. In this study, Cyanobacteria and Chlorophyta microalga groups images are captured by using an inverted microscope. Data augmentation process has been carried out to increase the classification success in Convolutional Neural Network (CNN) models. The collected images are classified by employing two different methods. For the first method, classification is performed with seven different CNN models. In the second method, the Support Vector Machine (SVM) is used to increase the classification success of the AlexNet model with the lowest accuracy. For this, deep features which are extracted from the AlexNet model are classified with SVM. Four different kernel functions are used in the SVM classification process. The highest accuracy is found to be 99.66% among the different CNN models. AlexNet, which has the lowest accuracy with 98%, has reached 99.66% accuracy as a result of its application with SVM.

Convolutional neural network - Support vector machine based approach for classification of cyanobacteria and chlorophyta microalgae groups

CNN–SVM hybrid model for varietal classification of wheat based on bulk samples

Wheat is a stable food raw material for the majority of people around the world as wheat-based products provide an important part of the daily energy intake in many countries Wheat is generally milled into flour prior to use in the bakery industry Flour yield is one of the major quality criteria in wheat milling Flout yield detection requires large amount of sample, costly machinery, longer time of tempering and milling practices requiring substantial workload In this study, artificial neural network (ANN) approach has been employed to predict flour milling yield The ANN was designed in the MATLAB software using such wheat physical properties as hectoliter weight, thousand-kernel weight, kernel size distribution, and grain hardness Flour yields and four different kernel physical features (hectoliter weight, thousand-kernel weight, kernel size distribution, and grain hardness) were first collected from 2400 wheat samples through the conventional methods The ANN was trained using 85% of 2400 flour yield data and tested with 15% of the remaining data In the training of the ANN, various models have been investigated to find the best ANN structure Additionally, two data sets with and without grain hardness have been employed to determine the effect of grain hardness on the prediction performance of the ANN model It was found that grain hardness which reduced the MAE values from 23333 to 22611 and RMSE values from 30775 to 29146 gave better result The results proved that the developed ANN model can be used to estimate flour yield using wheat kernel physical properties

https://www.ijisae.org/IJISAE/article/download/1084/619

Wheat Flour Milling Yield Estimation Based on Wheat Kernel Physical Properties Using Artificial Neural Networks

Rye translocation is a significant genetic resource used in the development of drought-tolerant wheat varieties. The effects of vernalization genes on root angles in the reciprocal recombinant inbred population carrying rye translocation were investigated in the study. Three hundred and five bread wheat lines obtained from reciprocal recombinant inbred line population, which were found to carry rye translocation by sodium dodecyl sulfate polyacrylamide gel electrophoresis and PCR analysis, were molecularly screened for Vrn genes, and their Vrn gene combinations were determined. The seeds of the lines were germinated in petri dishes in three repetitions, and the growth angles of the roots in the early period were determined by image processing technique. The mean root angles of the lines in reciprocal populations varied between 61° and 249°. The effect of Vrn allele combinations on root angles is statistically significant. Lines carrying the dominant Vrn-D1 allele affecting the spring growth character and the gene combinations with the recessive vrn-B1 allele affecting the winter growth character generally have narrower root angles. Similarly, the dominant Vrn-A1b allele was found to be particularly effective on the narrower root angle alone. In the Tahirova-2000/Tosunbey population, the allele combination associated with the narrowest root angle was Vrn-B1 x Vrn-D1 x Vrn-A1b. The results showed that the reciprocal population carrying the 1BL.1RS rye translocation showed a phenotypic variation for root angles, and Vrn genes affected the root angles. The developed population can be used, especially in drought studies, to develop elite material carrying the rye translocation.

Mesut Ersin Sonmez

Papers

Convolutional neural network - Support vector machine based approach for classification of cyanobacteria and chlorophyta microalgae groups

CNN–SVM hybrid model for varietal classification of wheat based on bulk samples

Wheat Flour Milling Yield Estimation Based on Wheat Kernel Physical Properties Using Artificial Neural Networks

Effect of vernalization (Vrn) genes on root angles of bread wheat lines carrying rye translocation

Deep learning-based classification of microalgae using light and scanning electron microscopy images.