[26]

2020

Mango sorting and grading according to their quality.

SqueezeNet deep CNN.

Nearly 10.000 mango fruits self-collected using smart phone, normal camera and a SEEK Thermal camera.

RGB images: 93.33%; Thermal images: 92.27%

[15]

2021

Metection of mangoes infected with anthracnose.

proposed CNN based on AlexNet architecture.

Dataset of images captured by a CDD camera in real conditions.

70%

[9]

2021

Segment and identify the diseases (Anthracnose and apical nacrosis) on the mango leave.

CNN based Fully-convolutional- network (FrCNnet) model.

Real-time self-collected

2286 images captured using different types of image capturing gadgets.

Accuracy: 99.2%; FNR: 0.8%

[24]

2021

Mango and grape diseases detection and identification.

AlexNet CNN.

7,222 and 1,266 images of diseased and healthy mango and grape leaves collected from the PlantVillage dataset and acquired locally respectively.

Grape: 99%

Mango: 89%

[17]

2022

Mango leaf disease identification and classification.

CNN with crossover-based levy flight distribution, MobileNetV2 model and SVM.

380 mango images self-captured from a mango cultivating land in India.

Not specified