Laser & Optoelectronics Progress, Volume. 56, Issue 23, 231007(2019)
Object Detection Model Based on Multi-Scale Feature Integration
Figures & Tables(15)
Fig. 8. Precision-Recall curves of two models. (a)(c)(e) YOLOv2 model;(b)(d)(f) RF-YOLOv2 model
Fig. 9. Detection results. (a)(c)(e)(g)(i) Detection results of YOLOv2 model; (b)(d)(f)(h)(j) detection results of RF-YOLOv2 model
Table 1. RF-YOLOv2 network structure
View tableTable 1. RF-YOLOv2 network structure
Layer block Type Number of filters Size / stride Output Convolutional 32 3×3 416×416 Maxpool 2×2/2 208×208 Convolutional 64 3×3 208×208 1× Convolutional 32 1×1 Convolutional 64 3×3 Residual 208×208 Maxpool 2×2/2 104×104 Convolutional 128 3×3 104×104 2× Convolutional 64 1×1 Convolutional 128 3×3 Residual 104×104 Maxpool 2×2/2 52×52 Convolutional 256 3×3 52×52 4× Convolutional 128 1×1 Convolutional 256 3×3 Residual 52×52 Maxpool 2×2/2 26×26 Convolutional 512 3×3 26×26 4× Convolutional 256 1×1 Convolutional 512 3×3 Residual 26×26 Maxpool 2×2/2 13×13 Convolutional 1024 3×3 13×13 4× Convolutional 512 1×1 Convolutional 1024 3×3 Residual 13×13 Avgpool Global 3 Softmax
Table 2. Comparison of accuracy and detection speed
View tableTable 2. Comparison of accuracy and detection speed
Model Accuracy of car /% Accuracy of pedestrian / % Accuracy of cyclist / % Detection speed / (frame·s-1) YOLOv2 68.56 44.26 55.95 46.4 RF- YOLOv2 87.88 52.91 74.05 30.3 YOLOv3 89.34 60.93 83.94 23.1
Table 3. Change process of recall rate and
I OUView tableTable 3. Change process of recall rate and
I OUNumber of training RF-YOLOv2 model YOLOv2 model Recall rate /% I OU /%Recall rate /% I OU /%10000 50.36 43.29 48.18 43.42 20000 55.45 46.34 53.11 45.98 30000 61.47 50.65 55.83 47.79 40000 64.92 52.56 54.13 46.72 50000 65.87 53.63 57.98 49.04
Table 4. Three sample detection results of car category
View tableTable 4. Three sample detection results of car category
Model Accuracy of easy sample /% Accuracy of moderate sample /% Accuracy of hard sample /% YOLOv2 70.56 57.32 50.44 Faster-rcnn 87.90 79.11 70.19 RF-YOLOv2 91.01 81.26 72.41
Table 5. Three sample detection results of pedestrian category
View tableTable 5. Three sample detection results of pedestrian category
Model Accuracy of easy sample /% Accuracy of moderate sample /% Accuracy of hard sample /% YOLOv2 59.97 49.05 44.91 Faster-rcnn 78.35 65.91 61.19 RF-YOLOv2 64.35 57.02 53.94
Table 6. Three sample detection results of cyclist category
View tableTable 6. Three sample detection results of cyclist category
Model Accuracy of easy sample /% Accuracy of moderate sample /% Accuracy of hard sample /% YOLOv2 56.47 56.68 53.02 Faster-rcnn 71.41 62.81 55.44 RF-YOLOv2 79.76 74.68 72.41
Tools
Get Citation
Copy Citation Text
Wanjun Liu, Feng Wang, Haicheng Qu. Object Detection Model Based on Multi-Scale Feature Integration[J]. Laser & Optoelectronics Progress, 2019, 56(23): 231007
Paper Information
Category: Image Processing
Received: May. 10, 2019
Accepted: Jun. 3, 2019
Published Online: Nov. 27, 2019
The Author Email: Feng Wang (838808390@qq.com)
© Copyright 2018-2021 | Chinese Laser Press.
All Rights Reserved 沪ICP备15018463号-20