ABR-fractional image enhancement for low-contrast grape leaf disease classification using GrNet-18 CNNs and genetic algorithm

Key Points

• Improved classification accuracy of low-contrast grape leaf diseases boosts diagnostic capabilities, showcasing the effectiveness of image enhancement methods.
• Utilization of the GrNet-18 convolutional neural network achieves significant advancements in recognizing subtle disease cues within grape leaves.
• Assessment implements a genetic algorithm to optimize parameters, enhancing the CNN's ability to classify low-contrast images effectively.
• Findings support the importance of advanced imaging techniques in agriculture, suggesting these methods may aid in earlier disease detection.