Indoor Scene Object Detection Based on Improved YOLOv4 Algorithm

Weigang Li; Chao Yang; Lin Jiang; Yuntao Zhao

doi:10.3788/LOP202259.1815003

Laser & Optoelectronics Progress, Volume. 59, Issue 18, 1815003(2022)

Indoor Scene Object Detection Based on Improved YOLOv4 Algorithm

Weigang Li^*, Chao Yang, Lin Jiang, and Yuntao Zhao

Engineering Research Center for Metallurgical Automation and Measurement Technology, Ministry of Education, Wuhan University of Science and Technology, Wuhan 430081, Hubei , China

show less

Abstract Get PDF(in Chinese)

Figures & Tables(14)

Fig. 1. Diagram of YOLOv4 network structure

Download full size

Fig. 2. Prediction box of YOLOv4 algorithm in 19×19 cells

Download full size

Fig. 3. Diagram of improved YOLOv4 network structure

Download full size

Fig. 4. CSPNet structured in SPP module

Download full size

Fig. 5. CSPNet structured in continuous convolution module

Download full size

Fig. 6. Depthwise separable convolution

Download full size

Fig. 7. Scatter plot of the size distribution of sample ground truth box and prior box

Download full size

Fig. 8. Accuracy comparison between K-means algorithm and K-means++ algorithm under original YOLOv4

Download full size

Fig. 9. Comparison of detection results before and after algorithm improvement

Download full size

Fig. 10. Comparison of detection results on each class before and after algorithm improvement

Download full size

Table 1. Performance results of the neck network integrated into the CSPNet structure
View table
Table 1. Performance results of the neck network integrated into the CSPNet structure
Method Neck module Parameters / $10^{7}$ mAP /% Speed /（frame·s^-1）
1 SPP+PAN 5.35 81.0 66.1
2 SPP+CSPPAN 4.65 81.8 66.5
3 CSPSPP+PAN 5.95 82.6 63.6
4 CSPSPP+CSPPAN 5.25 83.6 62.3

Table 2. Impact of depthwise separable convolution on network performance

View table

Table 2. Impact of depthwise separable convolution on network performance

Method	Backbone	Neck	Parameters/ $10^{7}$	mAP /%	Speed /（frame·s^-1）
1	CSPDarknet53	PAN	5.95	82.6	63.5
2	CSPDarknet53	CSPPAN	5.25	83.6	62.3
3	DS-CSPDarknet53	PAN	4.41	81.9	67.0
4	DS-CSPDarknet53	CSPPAN	3.89	83.2	68.0
5	CSPDarknet53	DS-PAN	4.72	82.0	68.7
6	CSPDarknet53	DS-CSPPAN	4.60	82.9	65.8
7	DS-CSPDarknet53	DS-PAN	3.28	81.9	72.9
8	DS-CSPDarknet53	DS-CSPPAN	3.24	83.0	72.1

Table 3. Performance comparison of different improvements

View table

Table 3. Performance comparison of different improvements

Improvement strategy			Weight size /MB	mAP /%	Speed /（frame·s^-1）
K-means++	CSP-Neck	DS	Weight size /MB	mAP /%	Speed /（frame·s^-1）
			102	79.8	66.1
		√	52	79.3	78.6
	√		101	81.2	62.4
	√	√	65	80.6	71.9
√			102	81.0	66.1
√		√	52	80.8	78.1
√	√		101	83.6	62.3
√	√	√	65	83.0	72.1

Table 4. Performance comparison of different detection algorithms
View table
Table 4. Performance comparison of different detection algorithms
Model Weight size /MB mAP /% Speed /（frame·s^-1）
Faster-RCNN-resnet50 315 79.0 21.3
SSD-resnet50 108 78.2 40.5
YOLOv3 118 75.2 60.8
YOLOv4 102 79.8 66.1
YOLOv5 94 81.5 72.5
Improved YOLOv4 65 83.0 72.1

Tools

Get Citation

Copy Citation Text

Weigang Li, Chao Yang, Lin Jiang, Yuntao Zhao. Indoor Scene Object Detection Based on Improved YOLOv4 Algorithm[J]. Laser & Optoelectronics Progress, 2022, 59(18): 1815003

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites