Automated Tomato Defect Detection Using CNN Feature Fusion for Enhanced Classification

Processes, Journal Year: 2025, Volume and Issue: 13(1), P. 115 - 115

Published: Jan. 4, 2025

Tomatoes are among the most widely cultivated and consumed vegetable crops worldwide. They usually harvested in large quantities that need to be promptly accurately classified into healthy defective categories. Traditional methods for tomato classification labor-intensive prone human error. Therefore, this study proposes an approach leverages feature fusion from two pre-trained convolutional neural networks (CNNs), VGG16 ResNet-50, enhance performance. A comprehensive evaluation of multiple individual hybrid classifiers was conducted on a dataset 43,843 images, which is heavily imbalanced toward class. The results showed best-performing classifier fused features achieved average precision (AP) accuracy 0.92 0.97, respectively, test set. In addition, experimental revealed improved performance across metrics, including accuracy, AP, recall, F1-score, compared ResNet-50. Furthermore, proposed benchmarked against three standalone CNN models, namely MobileNetV2, EfficientNetB0, DenseNet121, demonstrated superior all evaluated metrics. These findings highlight efficacy deep addressing class imbalance improving automated defect detection.

Language: Английский

---

Amending clayey and sandy soils with nano - bio phosphorous for regulating tomato growth, biochemical, and physiological characteristics

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

Published: Oct. 23, 2024

Phosphorus is a critical nutrient that significantly enhances tomato production, so maintaining an adequate level of phosphorus plays essential role in enhancing the growth by being present soil. This study assessed impact soil texture and content on plant properties using factorial, complete, randomized design with four replications. Treatments included clayey sandy soils varying sources: non-phosphorus (P0), calcium phosphate (CaP1 CaP2), nano-hydroxyapatite (PN1 PN2), where 1 indicates concentration 0.12 g 2 0.23 per 5-kilogram pot fertilizer. Results indicated treatments influenced yield parameters such as average fruit weight, juice content, antioxidant activity, volume. In soil, CaP2 treatment had superior effect yield, shoot fresh weight. comparison conditions, produced 50% increase number, 29% 91% yield. The then impacted weight root length, while appeared to be more dependent type than sources. Similar CaP1 treatments, PN1 clay also resulted highest dry weights shoots when compared control group. Generally, findings from this suggest use can serve reliable method improve growth, quality tomatoes, especially environments. However, nano-based phosphorous sources need tested see if they performance range conditions. Also, further research should look into long-term effects interventions health sustainability.

Language: Английский

---

Impact of cultivation methods and fertilization by EM (Effective microorganisms) and /or compost on Productivity and water use efficiency of wheat (Triticum aestivum L.)

Egyptian Journal of Soil Science, Journal Year: 2023, Volume and Issue: 64(1), P. 0 - 0

Published: Nov. 12, 2023

A field experiment was conducted at Mallawi Agriculture Research Station, Minia Governorate, Egypt for the two successive seasons of 2021/ 2022 and 2022/ 2023, to examine effect fertilization (100%NPK, 80 % NPK & EM applied (5L fed -1), compost 5 ton fed-1 Mixture compost) on wheat grown in clay soil under three cultivation methods, i.e. broad casting or ridges raised beds. The laid out a split plot design with replicates. Water productivity water use efficiency values were improved bed method compared other methods both seasons. highest averages plant height (cm). grain yield (ton fed-1), straw 1000 protein% seasons, detected from planting combined 80% NPK+ mix (EM compost). Nutrition contents total uptake kg increased same treatment.

Language: Английский

---

Metabolic modeling of a plant-pathogen interaction quantifies the metabolic bottlenecks underlying bacterial wilt

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2024, Volume and Issue: unknown

Published: Dec. 13, 2024

Abstract During plant infection, complex metabolic interactions take place between the and pathogen, among which occurs a true battle for resources. On one side, pathogen harvests nutrients at expense of to sustain its growth virulence. other tries prevent multiplication by competing or using anti-microbial compounds. Plants pathogens have thus contrary objectives in an intertwined system that is particularly difficult untangle experimentally. To help decipher this interaction, we used genome-scale modeling combining multi-organ model plant, quantitative data mathematical approach sequential flux balance analyses (FBAs). We applied modelling strategy interaction pathogenic bacterium Ralstonia pseudosolanacearum natural host tomato plant. This allowed, first time, fluxes matter occurring during infection. The showed that, pathosystem studied, i) plant’s photosynthetic capacity more stringent environmental condition than minerals bacterial proliferation ii) reduction transpiration what limits then stops later growth, iii) hijacking stem resources can boost but accessory, iv) pathogen-excreted putrescine predicted be directly reused biomass. study provides holistic view plant-pathogen highlighted criticality water flow when bacteria responsible infection fast-growing, xylem-colonizing one. Significance statement When infected two organisms. understand both metabolism. insight on consequences physiology Among results, decline growth.

Language: Английский

---