Table 1 Inferences from the literature review.
S. no. | Source | Methodology | Inference |
|---|---|---|---|
1 | Mu et al.2 | Mask R-CNN-based detection model followed by a king flower segmentation algorithm to identify and locate king flowers | Based on the flower stages of 20% to 80% blooming, the king flower detection accuracy varies from 98.7% to 65.6% |
2 | Almogdady et al.3 | Back-Propagation Artificial NN and chanvese active contour model for flower recognition | Flower classification performs with the accuracy of 81.19% |
3 | Chen et al.5 | DeepLab: Semantic segmentation with atrous convolution | Semantic segmentation task achieves 79.7 percent mIOU |
4 | Hocevaret et al.7 | FC estimation algorithm included HSL thresholding | 10% of erroneous executions were found with steady camera |
5 | Bargoti et al.8 | WS and CHT algorithms to detect and count individual apple fruits | Detection accuracy of apple with F1-score of 0.861 |
6 | Sun et al.9 | DeepLab-ResNet for detecting the fruit flowers | Achieves an average F1 score of 80.9% on the peach, pear and another apple datasets |
7 | Duman et al.10 | Deep Learning models for flower classification | ResNet performs with lower accuracy than other algorithms |
8 | Patel et al.20 | NAS-FPN and Faster-RCNN for flower classification | NAS-FPN and Faster-RCNN produces mAP score of 87.6% |
9 | Cibuk et al.21 | Deep CNN for flower species classification | Flower classification was done with the accuracy of 96.39% |
10 | Swati Kosankar et al.37 | MobileNet CNN for flower species classification | Flower classification is slightly compromised by time and space |
11 | Zhao et al.23 | Color constancy automatic network for flower classification | By reducing the interference of illumination factors on targets, CCAN has high accuracy |
12 | Yuan et al.24 | Multi-layer neural network for chrysanthemum recognition | Detects chrysanthemum flower with the accuracy of 95% |
13 | Touqeer Abbas et al.25 | Fast RCNN for flower species recognition | Recognizes the flower species with mAP score of 83.3% |
14 | Bae et al.27 | Multimodal CNN for flower classification | Accuracy found to be 10% higher than Recurrent NN |
15 | Mesut Togacar et al.38 | Feature selection with CNN for flower classification | SVM classifies with the accuracy of 98% |
16 | Mohanty et al.29 | GLCM and GA for flower classification | Potential association rules can be mined efficiently by GLCM |
17 | Dias et al.30 | Deep CNN for apple flower detection | Detects the apple flower with 90% accuracy |
18 | Zhou et al.31 | LGM based CNN for multi-class fruit blossom detection | Detects the fruit blossom with the mean precision of 74.33% |
19 | Shang et al.32 | ShuffleNetv2-Ghost model for detection of apple flowers | Detects the apple flowers with the mean precision of 88.40% |
20 | Sun et al.39 | DeepLab-ResNet for apple, peach and pear flower detection | Detects the apple flowers with F1 Score of 89.6% |
21 | Estrada et al.40 | Yolo multi-column deep NN for flower detection | Detect flowers with the MAE of 39.13, RMSE of 69.69 |
22 | Mu et al.34 | Mask R-CNN for detection of apple flowers | The king flower detection accuracy was 65.6% |
23 | Zhang et al.36 | Xception CNN for apple flower detection | Classifies the flower with 85% accuracy |
