Table 1 A summary of the recent works related to disease detection.
Work | Type of data (leaf) | Objective | Method | Perform. |
|---|---|---|---|---|
Zhang et al. (2017)23 | Diseased cucumber leaf | Lesion segmentation | SLIC, PHOG and EM | \(91.18\%\) |
Pankaja et al. (2018)17 | Flavia dataset | Leaf recognition and classification | Chebyshev moments and roundness, HSV color moments, DWT texture, SVM | \(96.29\%\) |
Sengar et al. (2018)26 | Diseased cherry leaf | Segmentation and quantization of lesion area | Adaptive intensity based thresholding | \(99.0\%\) |
Islam et al. (2019)18 | Diseased rice plant leaf | Rice disease classification | Two level DWT, Ensemble of linear classifier | \(95.0\%\) |
Singh (2019)25 | Diseased sunflower leaf | Segmentation and disease classification | PSO and texture feature, Minimum distance classifier | \(98.0\%\) |
Tampinongkol et al. (2020)19 | Diseased Jabon leaf | Classification of spots and blight disease | Three level DWT, SVM | \(84.67\%\) |
Dhingra et al. (2020)20 | Basil, Tomato, Cherry, Pepper, and Apple Leaf | Healthy and disease classification | DWT, Gabor filter, and Histogram binning pattern, Gaussian classifier | \(98.86\%\) |
Khan et al. (2020)24 | PlantVillage data set and real world data set | Disease classification | SLIC, GLCM and PHOG, Random Forest | \(93.12\%\) |
Chouhan et al. (2021)33 | Jatropha Curcas and Pongamia Pinnata | Segmentation and classification | SLIC, ADALINE, SIFT and LBP | \(98.5\%\) |
Pandey et al. (2021)27 | Diseased vigna mungo leaves | Healthy and disease level identification | Statistical and GLCM features, PCA, SVM | \(95.69\%\) |
Chakraborty et al. (2022)28 | Potato leaves | Healthy and disease identification | Improved VGG16 | \(97.89\%\) |
Phan et al. (2022)29 | Corn leaves | Healthy and disease identification | SLIC and pre-trained model | \(97.77\%\) |
Abisha et al. (2023)30 | Diseased brinjal leaves | Healthy and disease identification | Deep CNN, Discrete Shearlet transform | \(93.30\%\) |
Singla et al. (2024)31 | PlantVillage data set | Healthy and disease identification | MobileNet | \(97.35\%\) |
Bedi et al. (2024)32 | PlantVillage data set | Disease identification and severity estimation | Few-Shot Learning | \(99.00\%\) |
Proposed method | Pongamia pinnata leaf | Disease severity finding | Optimized SLIC and proposed MSCS based feature selection | \(97.60\%\) |
