Recognition performance of different artificial neural networks for distinguishing banana slices subjected to different combinations of pretreatment and microwave drying

Food Control, Journal Year: 2024, Volume and Issue: 163, P. 110488 - 110488

Published: March 31, 2024

Language: Английский

---

Non-destructive discrimination of vacuum-dried banana using image processing operation and machine learning approach

Food and Bioproducts Processing, Journal Year: 2023, Volume and Issue: 141, P. 36 - 48

Published: July 11, 2023

Language: Английский

---

Exploration of Convective and Infrared Drying Effect on Image Texture Parameters of ‘Mejhoul’ and ‘Boufeggous’ Date Palm Fruit Using Machine Learning Models

Foods, Journal Year: 2024, Volume and Issue: 13(11), P. 1602 - 1602

Published: May 21, 2024

Date palm (Phoenix dactylifera L.) fruit samples belonging to the ‘Mejhoul’ and ‘Boufeggous’ cultivars were harvested at Tamar stage used in our experiments. Before scanning, date dried using convective drying 60 °C infrared with a frequency of 50 Hz, then they scanned. The scanning trials performed for two hundred fresh, convective-dried, infrared-dried forms each cultivar flatbed scanner. image-texture parameters extracted from images converted individual color channels RGB, Lab, XYZ, UVS models. models classify fresh developed based on selected image textures machine learning algorithms groups Bayes, Trees, Lazy, Functions, Meta. For both cultivars, built Random Forest group Trees turned out be accurate successful. average classification accuracy reached 99.33%, whereas distinguished an 94.33%. In case model, higher correctness discrimination was between samples, highest number misclassified cases occurred convective-dried fruit. Thus, procedure may considered innovative approach non-destructive assessment impact external quality characteristics

Language: Английский

---

The Changes in Bell Pepper Flesh as a Result of Lacto-Fermentation Evaluated Using Image Features and Machine Learning

Foods, Journal Year: 2022, Volume and Issue: 11(19), P. 2956 - 2956

Published: Sept. 21, 2022

Food processing allows for maintaining the quality of perishable products and extending their shelf life. Nondestructive procedures combining image analysis machine learning can be used to control processed foods. This study was aimed at developing an innovative approach distinguishing fresh lacto-fermented red bell pepper samples involving selected textures algorithms. Before processing, pieces subjected spontaneous lacto-fermentation were imaged using a digital camera. The texture parameters extracted from images converted different color channels

Language: Английский

---

Rapid Detection of Changes in Image Textures of Carrots Caused by Freeze-Drying using Image Processing Techniques and Machine Learning Algorithms

Sustainability, Journal Year: 2023, Volume and Issue: 15(8), P. 7011 - 7011

Published: April 21, 2023

The objective of this study was to evaluate the differences in texture parameters between freeze-dried and fresh carrot slices using image processing artificial intelligence. Images were acquired a digital camera. Texture extracted from slice images converted individual color channels L, a, b, R, G, B, X, Y, Z. A total 1629 parameters, 181 for each these channels, obtained. Models classification created various machine learning algorithms, based on attributes selected combined set textures all (L, Z). Using three different feature selection methods (Genetic Search, Ranker, Best First), 20 most effective determined method. models with highest accuracy obtained by applying algorithms Trees, Rules, Meta, Lazy, Functions groups determined. successes compared. Random Forest, Multi-class Classifier, Logistic SMO achieved 100% performed features algorithm.

Language: Английский

---

Application of Imaging and Artificial Intelligence for Quality Monitoring of Stored Black Currant (Ribes nigrum L.)

Foods, Journal Year: 2022, Volume and Issue: 11(22), P. 3589 - 3589

Published: Nov. 11, 2022

The objective of this study was to assess the influence storage under different conditions on black currant quality in a non-destructive and inexpensive manner using image processing artificial intelligence. Black currants were stored at room temperature 20 ± 1 °C 3 (refrigerator). images directly after harvest fruit for one two weeks obtained digital camera. Then, texture parameters computed from converted color channels R (red), G (green), B (blue), L (lightness component white), (green negative red positive values), b (blue yellow X (component with information), Y (lightness), Z information). Models classification built various machine learning algorithms based selected textures RGB, Lab, XYZ spaces. IBk, multilayer perceptron, multiclass classifier RGB space, IBk algorithm Lab space distinguished unstored samples an average accuracy 100%, kappa statistic weighted averages precision, recall, Matthews correlation coefficient (MCC), receiver operating characteristic (ROC) area, precision-recall (PRC) area equal 1.000. This indicated very distinct change external structure first week more visible changes increasing time. A reaching 98.67% (multilayer space) refrigerator may indicate smaller caused by low temperature. approach combining intelligence turned out be promising monitor during storage.

Language: Английский

---

Innovative Models Built Based on Image Textures Using Traditional Machine Learning Algorithms for Distinguishing Different Varieties of Moroccan Date Palm Fruit (Phoenix dactylifera L.)

Agriculture, Journal Year: 2022, Volume and Issue: 13(1), P. 26 - 26

Published: Dec. 22, 2022

The aim of this study was to develop the procedure for varietal discrimination date palm fruit using image analysis and traditional machine learning techniques. images ‘Mejhoul’, ‘Boufeggous’, ‘Aziza’, ‘Assiane’, ‘Bousthammi’ varieties, converted individual color channels, were processed extract texture parameters. After performing attribute selection, textures used build models intended different varieties algorithms from Functions, Bayes, Lazy, Meta, Trees groups. Models developed combining selected a set all channels sets spaces channels. models, including combined distinguished five with an average accuracy reaching 98%, ‘Mejhoul’ completely correctly discriminated SMO (Functions) IBk (Lazy) algorithms. By reducing number correctness classification increased. three most ‘Bousthammi’, revealed 100% each algorithm (SMO, Naive Bayes (Bayes), IBk, LogitBoost (Meta), LMT (Trees)). In case accuracies lower, 97.3% space RGB 99.63% fruits. results can be in practice vision systems sorting distinguishing authenticate variety further processing.

Language: Английский

---

Machine learning and computer vision technology to analyze and discriminate soil samples

Scientific Reports, Journal Year: 2024, Volume and Issue: 14(1)

Published: Aug. 27, 2024

Soil texture is one of the most important elements to consider before planting and tillage. These features affect product selection regulate its water permeability. Discrimination soils by determining soil requires an intense workload time-consuming. Therefore, having a powerful tool knowledge for texture-based discrimination could enable rapid accurate soils. This study focuses on presenting new models 6 different sample groups (Soil_1 Soil_6) based 12 machine learning algorithms that can be utilized various problems. As result, overall accuracy values were determined as greater than 99.2% (Trilayered Neural Network). The greatest value was found in Bayes Net (99.83%) followed Subspace Discriminant (99.80%). In algorithm, MCC (Matthews Correlation Coefficient) F-measure obtained 0.994 0.995 Soil_4 Soil_6 while these 1.000 other groups. types visually vary their texture, mineral composition, moisture levels. variability this influenced fertilization, precipitation levels, cultivation. It capture images conditions are more stable. conclusion, present has proven feasibility rapid, non-destructive, image processing-based learning.

Language: Английский

---

Non-Destructive Monitoring of Sweet Pepper Samples After Selected Periods of Lacto-Fermentation

Agriculture, Journal Year: 2024, Volume and Issue: 14(11), P. 1855 - 1855

Published: Oct. 22, 2024

Fermented food is characterized by positive health-promoting properties. The objective of this study was to distinguish and assess the changes in flesh structure sweet bell pepper samples after specific periods fermentation a non-destructive manner. Two cultivars pepper, red yellow, were subjected lacto-fermentation. experiments lasted 56 days taken for analysis at beginning (0 days) 3, 7, 10, 14, 21, 28, days. process monitored based on image features, which used develop machine learning models distinguishing before various lacto-fermentation (0, days). average accuracy classification up 93% model built using IBk (Lazy group). yellow distinguished 90% LMT algorithm (Trees performed allowed us determine terms textures during

Language: Английский

---

The innovative approach to the assessment of differences in image textures between windfall apple samples dried using non-thermal and thermal techniques without and with ultrasound pretreatment

Expert Systems with Applications, Journal Year: 2023, Volume and Issue: 232, P. 120917 - 120917

Published: June 28, 2023

Language: Английский

---