Figure 1.
The flowchart of the proposed algorithm.
Figure 1.
The flowchart of the proposed algorithm.
Figure 2.
P-nearest neighbors evenly spaced from the center pixel and circle of radius Rs.
Figure 2.
P-nearest neighbors evenly spaced from the center pixel and circle of radius Rs.
Figure 3.
Collection equipment of the point cloud scene. (a) Backpack mobile surveying and mapping robots and (b) Matterport camera.
Figure 3.
Collection equipment of the point cloud scene. (a) Backpack mobile surveying and mapping robots and (b) Matterport camera.
Figure 4.
Scene 1 point cloud data. (a) Original point cloud of the training set; (b) ground truth for the training set; (c) original point cloud of the test set; and (d) ground truth for the test set. Red, yellow, and blue points represent trees, cars, and floors, respectively.
Figure 4.
Scene 1 point cloud data. (a) Original point cloud of the training set; (b) ground truth for the training set; (c) original point cloud of the test set; and (d) ground truth for the test set. Red, yellow, and blue points represent trees, cars, and floors, respectively.
Figure 5.
Scene 2 point cloud data. (a) Original point cloud of the training set; (b) ground truth of the training set; (c) original point cloud of the test set; and (d) ground truth of the test set. Red, green, yellow, and blue points represent trees, cars, buildings, and floors, respectively.
Figure 5.
Scene 2 point cloud data. (a) Original point cloud of the training set; (b) ground truth of the training set; (c) original point cloud of the test set; and (d) ground truth of the test set. Red, green, yellow, and blue points represent trees, cars, buildings, and floors, respectively.
Figure 6.
Scene 3 point cloud data. (a) Original point cloud of the training set; (b) ground truth of the training set; (c) original point cloud of the test set; and (d) ground truth of the test set. Red, green, yellow, and blue points represent trees, cars, buildings, and floors, respectively.
Figure 6.
Scene 3 point cloud data. (a) Original point cloud of the training set; (b) ground truth of the training set; (c) original point cloud of the test set; and (d) ground truth of the test set. Red, green, yellow, and blue points represent trees, cars, buildings, and floors, respectively.
Figure 7.
Scene 4 point cloud data. (a) Original point cloud of the training set; (b) ground truth of the training set; (c) original point cloud of the test set; and (d) ground truth of the test set. Red, green, and blue points represent the table, floor, and chair, respectively.
Figure 7.
Scene 4 point cloud data. (a) Original point cloud of the training set; (b) ground truth of the training set; (c) original point cloud of the test set; and (d) ground truth of the test set. Red, green, and blue points represent the table, floor, and chair, respectively.
Figure 8.
Scene 5 point cloud data. (a) Original point cloud of the training set; (b) ground truth of the training set; (c) original point cloud of the test set; and (d) ground truth of the test set. Red, green, yellow, and blue points represent plants, tables, floors, and chairs, respectively.
Figure 8.
Scene 5 point cloud data. (a) Original point cloud of the training set; (b) ground truth of the training set; (c) original point cloud of the test set; and (d) ground truth of the test set. Red, green, yellow, and blue points represent plants, tables, floors, and chairs, respectively.
Figure 9.
Classification results for Scene 1. (a) Ground truth; (b) Method 1; (c) Method 2; (d) Method 3; and (e) our method. Red, blue, and yellow points represent trees, floors, and cars, respectively.
Figure 9.
Classification results for Scene 1. (a) Ground truth; (b) Method 1; (c) Method 2; (d) Method 3; and (e) our method. Red, blue, and yellow points represent trees, floors, and cars, respectively.
Figure 10.
Classification results for Scene 2. (a) Ground truth; (b) Method 1; (c) Method 2; (d) Method 3; and (e) our method. Red, green, blue, and yellow points represent trees, buildings, floors, and cars, respectively.
Figure 10.
Classification results for Scene 2. (a) Ground truth; (b) Method 1; (c) Method 2; (d) Method 3; and (e) our method. Red, green, blue, and yellow points represent trees, buildings, floors, and cars, respectively.
Figure 11.
Classification results for Scene 3. (a) Ground truth; (b) Method 1; (c) Method 2; (d) Method 3; and (e) our method. Red, green, blue, and yellow points represent trees, buildings, floors, and cars, respectively.
Figure 11.
Classification results for Scene 3. (a) Ground truth; (b) Method 1; (c) Method 2; (d) Method 3; and (e) our method. Red, green, blue, and yellow points represent trees, buildings, floors, and cars, respectively.
Figure 12.
Classification results for Scene 4. (a) Ground truth; (b) Method 1; (c) Method 2; (d) Method 3; and (e) our method. Red, green, and blue points represent the table, floor, and chair, respectively.
Figure 12.
Classification results for Scene 4. (a) Ground truth; (b) Method 1; (c) Method 2; (d) Method 3; and (e) our method. Red, green, and blue points represent the table, floor, and chair, respectively.
Figure 13.
Classification results for Scene 5. (a) Ground truth; (b) Method 1; (c) Method 2; (d) Method 3; and (e) our method. Red, green, blue, and yellow points represent flowers, tables, chairs, and floors, respectively.
Figure 13.
Classification results for Scene 5. (a) Ground truth; (b) Method 1; (c) Method 2; (d) Method 3; and (e) our method. Red, green, blue, and yellow points represent flowers, tables, chairs, and floors, respectively.
Table 1.
Statistics of experimental datasets.
Table 1.
Statistics of experimental datasets.
| Train | Test |
---|
| Floor | Building | Car | Tree | Floor | Building | Car | Tree |
Scene 1 | 54.327 | | 29.523 | 46.068 | 25.038 | | 63.174 | 93.852 |
Scene 2 | 71.587 | 123.521 | 10.545 | 28.754 | 29.080 | 46.854 | 5918 | 7248 |
Scene 3 | 119.255 | 180.919 | 15.394 | 13.504 | 155.931 | 201.930 | 17.492 | 87.601 |
| Chair | Table | Floor | Flower | Chair | Table | Floor | Flower |
Scene 4 | 24.025 | 17.671 | 77.880 | | 24.713 | 29.467 | 74.692 | |
Scene 5 | 30.801 | 16.148 | 62.596 | 18.394 | 28.748 | 14.611 | 47.662 | 19.278 |
Table 2.
Definition relationships between predicted and true values.
Table 2.
Definition relationships between predicted and true values.
Ground Truth | Predicted |
---|
Positive | Negative |
---|
Positive | True Positive (Tp) | False Negative (Fn) |
Negative | False Positive (Fp) | True Negative (Tn) |
Table 3.
Comparison of the Kappa/OA value (%) of each point cloud scene under different classifiers and feature conditions. For each scene, bold font indicates the best result of each classifier.
Table 3.
Comparison of the Kappa/OA value (%) of each point cloud scene under different classifiers and feature conditions. For each scene, bold font indicates the best result of each classifier.
Classifiers | Features | Scene 1 | Scene 2 | Scene 3 | Scene 4 | Scene 5 |
---|
RF | | 51.9/73.8 | 32.0/53.6 | 33.8/55.7 | 43.3/71.0 | 45.7/65.8 |
| 84.0/90.7 | 79.1/87.6 | 73.0/82.9 | 81.9/90.9 | 82.5/88.0 |
| 85.7/91.4 | 79.1/87.7 | 73.6/83.3 | 91.4/95.9 | 84.4/88.1 |
| 86.8/92.1 | 79.4/87.9 | 73.2/83.1 | 94.4/97.0 | 84.6/89.5 |
MLP | | 56.7/75.7 | 32.1/52.5 | 36.7/55.5 | 42.4/69.5 | 36.8/51.2 |
| 79.0/86.4 | 60.8/73.5 | 57.4/73.3 | 72.9/82.8 | 78.2/84.5 |
| 84.9/90.5 | 51.4/69.5 | 60.8/71.8 | 74.0/84.9 | 78.0/85.1 |
| 85.8/91.4 | 63.5/78.4 | 57.7/73.7 | 88.1/94.0 | 81.1/86.0 |
SVM | | 57.0/75.6 | 34.1/52.4 | 32.8/43.8 | 35.6/70.2 | 36.8/52.8 |
| 80.0/88.0 | 53.9/73.3 | 52.0/70.0 | 74.5/85.7 | 79.2/85.7 |
| 84.5/91.1 | 54.2/73.4 | 53.1/70.5 | 77.1/84.9 | 79.5/87.3 |
| 85.0/90.7 | 55.6/74.2 | 56.5/73.5 | 90.8/95.3 | 83.5/88.6 |
PointNet | Deep feature | 62.1/76.3 | 54.2/61.4 | 57.7/74.9 | 58.7/82.9 | 50.8/61.2 |
Table 4.
Statistics of average running time under the condition of different features and classifiers.
Table 4.
Statistics of average running time under the condition of different features and classifiers.
Feature | Average Time (min) |
---|
FVLBP | 2.86 |
FRGB | 0.58 |
FNormal | 1.73 |
FFPFH | 1.47 |
Classifier | Average Time (min) |
RF | 3.80 |
MLP | 5.03 |
SVM | 16.3 |
PointNet | 145.86 |
Table 5.
Comparison with the four methods in terms of “Feature Expression”.
Table 5.
Comparison with the four methods in terms of “Feature Expression”.
Method | Feature | Dimension |
---|
Method 1 | FVLBP | 10 |
Method 2 | FNormal + FFPFH | 36 |
Method 3 | FNormal + FFPFH + FRGB | 39 |
Our Method | FNormal + FFPFH + FRGB + FVLBP | 49 |
Table 6.
Comparison of the classification effects of ground objects and ground objects in each scene with precision/recall/F1-scores. For each scene, bold font indicates the best results of each metric.
Table 6.
Comparison of the classification effects of ground objects and ground objects in each scene with precision/recall/F1-scores. For each scene, bold font indicates the best results of each metric.
| Method | Floor | Car | Tree | Building | Kappa (%) | OA (%) |
---|
Scene 1 | Method 1 | 0.71/0.71/0.71 | 0.27/0.12/0.17 | 0.80/0.92/0.86 | | 51.9 | 73.8 |
Method 2 | 0.96/0.81/0.88 | 0.68/0.84/0.75 | 0.95/0.99/0.97 | | 84.0 | 90.7 |
Method 3 | 0.96/0.82/0.88 | 0.69/0.85/0.76 | 0.96/0.99/0.98 | | 85.7 | 91.4 |
Our | 0.96/0.83/0.89 | 0.70/0.87/0.77 | 0.97/1.00/0.98 | | 86.8 | 92.1 |
Scene 2 | Method 1 | 0.47/0.82/0.60 | 0.22/0.06/0.09 | 0.39/0.31/0.35 | 0.68/0.46/0.60 | 32.0 | 53.6 |
Method 2 | 0.91/0.88/0.90 | 0.87/0.55/0.68 | 0.79/0.73/0.76 | 0.87/0.94/0.90 | 79.1 | 87.6 |
Method 3 | 0.91/0.88/0.90 | 0.89/0.55/0.68 | 0.80/0.74/0.77 | 0.87/0.94/0.90 | 79.1 | 87.7 |
Our | 0.91/0.89/0.90 | 0.89/0.57/0.69 | 0.82/0.75/0.77 | 0.87/0.94/0.90 | 79.4 | 87.9 |
Scene 3 | Method 1 | 0.71/0.51/0.59 | 0.11/0.02/0.03 | 0.60/0.06/0.12 | 0.51/0.86/0.64 | 33.8 | 55.7 |
Method 2 | 0.99/0.91/0.95 | 0.80/0.30/0.43 | 0.92/0.45/0.60 | 0.73/0.98/0.84 | 73.0 | 82.9 |
Method 3 | 0.99/0.91/0.95 | 0.45/0.28/0.34 | 0.93/0.53/0.68 | 0.74/0.95/0.83 | 73.6 | 83.3 |
Our | 0.99/0.91/0.95 | 0.58/0.29/0.38 | 0.92/0.48/0.63 | 0.73/0.96/0.83 | 73.2 | 83.1 |
| Method | Chair | Table | Floor | Flower | Kappa (%) | OA (%) |
Scene 4 | Method 1 | 0.30/0.20/0.24 | 0.72/0.37/0.49 | 0.78/0.95/0.86 | | 43.3 | 71.0 |
Method 2 | 0.89/0.99/0.94 | 0.69/0.80/0.74 | 0.98/0.91/0.94 | | 81.9 | 90.9 |
Method 3 | 0.90/0.99/0.94 | 0.90/0.89/0.89 | 0.99/0.97/0.98 | | 91.4 | 95.9 |
Our | 0.90/0.99/0.94 | 0.97/0.90/0.93 | 0.99/0.98/0.99 | | 94.4 | 97.1 |
Scene 5 | Method 1 | 0.53/0.33/0.41 | 0.67/0.55/0.60 | 0.69/0.87/0.77 | 0.66/0.71/0.68 | 45.7 | 65.8 |
Method 2 | 0.89/0.77/0.83 | 0.79/0.75/0.77 | 0.92/0.97/0.94 | 0.83/0.92/0.88 | 82.5 | 88.0 |
Method 3 | 0.90/0.78/0.83 | 0.82/0.80/0.82 | 0.94/0.98/0.96 | 0.82/0.92/0.88 | 84.4 | 88.1 |
Our | 0.90/0.78/0.83 | 0.83/0.80/0.82 | 0.94/0.98/0.96 | 0.84/0.93/0.88 | 84.6 | 89.5 |