Research Article
[Retracted] Robot Obstacle Recognition and Target Tracking Based on Binocular Vision
Table 5
Analysis of obstacle recognition performance of YOLO convolutional neural network.
| Group | Target category (from left to right) | Confidence level (%) | x | y | | h | Time-consuming (s) |
| (a) | Dog | 100 | 357 | 29 | 354 | 409 | 0.089 |
| (b) | Bicycle | 100 | 302 | 499 | 533 | 408 | 0.088 | Car | 98 | 3 | 210 | 191 | 70 |
| (c) | Car | 38 | 165 | 220 | 54 | 49 | 0.095 | Person | 99 | 342 | 29 | 121 | 321 |
| (d) | Car | 94 | 581 | 170 | 90 | 114 | 0.116 | Car | 90 | 0 | 114 | 67 | 87 | Car | 96 | 12 | 109 | 143 | 69 | Car | 99 | 325 | 127 | 332 | 238 | Car | 65 | 571 | 123 | 42 | 49 | Car | 99 | 587 | 117 | 176 | 156 |
|
|