Multi-Object Detection in Security Screening Scene Based on Convolutional Neural Network
The technique for target detection based on a convolutional neural network has been widely implemented in the industry. However, the detection accuracy of X-ray images in security screening scenarios still requires improvement. This paper proposes a coupled multi-scale feature extraction and multi-s...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-10-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/22/20/7836 |
| _version_ | 1797470042516881408 |
|---|---|
| author | Fan Sun Xiangfeng Zhang Yunzhong Liu Hong Jiang |
| author_facet | Fan Sun Xiangfeng Zhang Yunzhong Liu Hong Jiang |
| author_sort | Fan Sun |
| collection | DOAJ |
| description | The technique for target detection based on a convolutional neural network has been widely implemented in the industry. However, the detection accuracy of X-ray images in security screening scenarios still requires improvement. This paper proposes a coupled multi-scale feature extraction and multi-scale attention architecture. We integrate this architecture into the Single Shot MultiBox Detector (SSD) algorithm and find that it can significantly improve the effectiveness of target detection. Firstly, ResNet is used as the backbone network to replace the original VGG network to improve the feature extraction capability of the convolutional neural network for images. Secondly, a multi-scale feature extraction (MSE) structure is designed to enrich the information contained in the multi-stage prediction feature layer. Finally, the multi-scale attention architecture (MSA) is fused onto the prediction feature layer to eliminate the redundant features’ interference and extract effective contextual information. In addition, a combination of Adaptive-NMS and Soft-NMS is used to output the final prediction anchor boxes when performing non-maximum suppression. The results of the experiments show that the improved method improves the mean average precision (mAP) value by 7.4% compared to the original approach. New modules make detection much more accurate while keeping the detection speed the same. |
| first_indexed | 2024-03-09T19:31:11Z |
| format | Article |
| id | doaj.art-e24e7b89cf9046b7861ce0f66cc03d00 |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-09T19:31:11Z |
| publishDate | 2022-10-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-e24e7b89cf9046b7861ce0f66cc03d002023-11-24T02:27:01ZengMDPI AGSensors1424-82202022-10-012220783610.3390/s22207836Multi-Object Detection in Security Screening Scene Based on Convolutional Neural NetworkFan Sun0Xiangfeng Zhang1Yunzhong Liu2Hong Jiang3College of Intelligent Manufacturing and Industrial Modernization, Xinjiang University, Urumchi 830017, ChinaCollege of Intelligent Manufacturing and Industrial Modernization, Xinjiang University, Urumchi 830017, ChinaCollege of Intelligent Manufacturing and Industrial Modernization, Xinjiang University, Urumchi 830017, ChinaCollege of Intelligent Manufacturing and Industrial Modernization, Xinjiang University, Urumchi 830017, ChinaThe technique for target detection based on a convolutional neural network has been widely implemented in the industry. However, the detection accuracy of X-ray images in security screening scenarios still requires improvement. This paper proposes a coupled multi-scale feature extraction and multi-scale attention architecture. We integrate this architecture into the Single Shot MultiBox Detector (SSD) algorithm and find that it can significantly improve the effectiveness of target detection. Firstly, ResNet is used as the backbone network to replace the original VGG network to improve the feature extraction capability of the convolutional neural network for images. Secondly, a multi-scale feature extraction (MSE) structure is designed to enrich the information contained in the multi-stage prediction feature layer. Finally, the multi-scale attention architecture (MSA) is fused onto the prediction feature layer to eliminate the redundant features’ interference and extract effective contextual information. In addition, a combination of Adaptive-NMS and Soft-NMS is used to output the final prediction anchor boxes when performing non-maximum suppression. The results of the experiments show that the improved method improves the mean average precision (mAP) value by 7.4% compared to the original approach. New modules make detection much more accurate while keeping the detection speed the same.https://www.mdpi.com/1424-8220/22/20/7836security screening scenesconvolutional neural networksmulti-scale feature extractionattentional mechanisms |
| spellingShingle | Fan Sun Xiangfeng Zhang Yunzhong Liu Hong Jiang Multi-Object Detection in Security Screening Scene Based on Convolutional Neural Network Sensors security screening scenes convolutional neural networks multi-scale feature extraction attentional mechanisms |
| title | Multi-Object Detection in Security Screening Scene Based on Convolutional Neural Network |
| title_full | Multi-Object Detection in Security Screening Scene Based on Convolutional Neural Network |
| title_fullStr | Multi-Object Detection in Security Screening Scene Based on Convolutional Neural Network |
| title_full_unstemmed | Multi-Object Detection in Security Screening Scene Based on Convolutional Neural Network |
| title_short | Multi-Object Detection in Security Screening Scene Based on Convolutional Neural Network |
| title_sort | multi object detection in security screening scene based on convolutional neural network |
| topic | security screening scenes convolutional neural networks multi-scale feature extraction attentional mechanisms |
| url | https://www.mdpi.com/1424-8220/22/20/7836 |
| work_keys_str_mv | AT fansun multiobjectdetectioninsecurityscreeningscenebasedonconvolutionalneuralnetwork AT xiangfengzhang multiobjectdetectioninsecurityscreeningscenebasedonconvolutionalneuralnetwork AT yunzhongliu multiobjectdetectioninsecurityscreeningscenebasedonconvolutionalneuralnetwork AT hongjiang multiobjectdetectioninsecurityscreeningscenebasedonconvolutionalneuralnetwork |