Research Article
Study on Waste Type Identification Method Based on Bird Flock Neural Network
Table 2
Comparison of recognition results after 200 samples are supplemented online.
| | | Hidden layer neurons | Recognition accuracy | Time consumption (s) | | Paper (%) | Plastic (%) | Glass (%) | Cloth (%) | Nonrecyclables (%) | Total (%) |
| | BPNN | 10 | 74.8 | 47.3 | 44.9 | 56.9 | 72.0 | 60.1 | 1.12 | | GA-BPNN | 10 | 73.0 | 27.8 | 60.5 | 62.1 | 69.4 | 58.8 | 2.57 | | PSO-BPNN | 10 | 81.4 | 41.6 | 55.4 | 58.2 | 67.4 | 60.9 | 2.50 | | BFNN | 10 | 87.6 | 76.9 | 72.0 | 84.5 | 85.3 | 81.0 | 2.83 |
|
|
