Abstract:X-ray inspections of contraband are widely used to maintain public transportation safety and protect life and property when people travel. To improve detection accuracy and reduce the probability of missed and false detection, a contraband detection algorithm YOLOv8s-DCN-EMA-IPIO* based on YOLOv8s is proposed. Firstly, the super-resolution reconstruction method based on the SRGAN network enhances the original data set, which is more conducive to model training. Secondly, DCNv2 (deformable convolution net v2) is introduced in the backbone network and merged with the C2f layer to improve the ability of the feature extraction and robustness of the model. Then, an EMA (efficient multi-scale attention) mechanism is proposed to suppress the interference of complex background noise and occlusion overlap in the detection process. Finally, the IPIO (improved pigeon-inspired optimization), which is based on the cross-mutation strategy, is employed to maximize the convolutional neural network's learning rate to derive the optimal group's weight information and ultimately improve the model's detection and recognition accuracy. The experimental results show that on the self-built data set, the mAP (mean average precision) of the improved model YOLOv8s-DCN-EMA-IPIO* is 73.43%, 3.98% higher than that of the original model YOLOv8s, and the FPS is 95, meeting the deployment requirements of both high precision and real-time.

What problem does this paper attempt to address?

The paper attempts to address the issue of improving the accuracy and real-time performance of contraband detection in X-ray security images. Specifically, the authors propose a contraband detection scheme based on an improved YOLOv8s network model (YOLOv8s-DCN-EMA-IPIO*) to reduce the probability of missed and false detections. ### Main Issues 1. **Missed and False Detections**: Traditional detection methods are prone to missed and false detections when dealing with a large amount of luggage and complex background noise. 2. **Small Object Detection**: Contraband items are usually small in size, and ordinary network models perform poorly in detecting small objects. 3. **Complex Background Interference**: Background noise and occlusion overlap in X-ray images interfere with the detection process, reducing detection accuracy. ### Solutions To address the above issues, the authors propose the following improvements: 1. **Data Augmentation**: Enhance the original dataset using SRGAN super-resolution reconstruction technology to improve the resolution and brightness of the images, making the appearance and shape of the items to be detected more precise, which is beneficial for the model's feature extraction operations. 2. **Introduction of Deformable Convolutional Networks (DCNv2)**: Introduce DCNv2 into the backbone network and merge it with the C2f layer to improve the model's feature extraction capability and robustness. 3. **Propose Efficient Multi-Scale Attention Mechanism (EMA)**: Suppress the interference of complex background noise and occlusion overlap during detection, increasing the model's focus on the target area. 4. **Optimize Learning Rate Parameters**: Use the pigeon-inspired optimization algorithm (IPIO) based on a cross-mutation strategy to generate the initial learning rate of the model by simulating the homing behavior of pigeons, and obtain the optimal learning rate through continuous iteration, ultimately improving the accuracy of target detection. ### Experimental Results Experimental results show that on the self-built dataset, the improved model YOLOv8s-DCN-EMA-IPIO* achieves a mAP (mean Average Precision) of 73.43%, which is 3.98% higher than the original YOLOv8s model, and the FPS (Frames Per Second) reaches 95, meeting the deployment requirements for high accuracy and real-time performance. ### Conclusion Through the above improvements, this paper successfully enhances the accuracy and real-time performance of contraband detection in X-ray security images, providing strong technical support for the safety of public transportation.

A Contraband Detection Scheme in X-ray Security Images Based on Improved YOLOv8s Network Model

Research on X-ray security contraband identification technology based on lightweight YOLOv8

Contraband detection algorithm for X‐ray security inspection images based on global semantic enhancement

X-YOLO: An Efficient Detection Network of Dangerous Objects in X-Ray Baggage Images

SC-YOLOv8: A Security Check Model for the Inspection of Prohibited Items in X-ray Images

X-Ray Small Target Security Inspection Based on TB-YOLOv5

Improved YOLOv8 Detection Algorithm in Security Inspection Image

Lightweight Detection Method for X-ray Security Inspection with Occlusion

Fine-YOLO: A Simplified X-ray Prohibited Object Detection Network Based on Feature Aggregation and Normalized Wasserstein Distance

Detection Algorithm of Laboratory Personnel Irregularities Based on Improved YOLOv7

Object Detection for Remote Sensing Based on the Enhanced YOLOv8 With WBiFPN

EBSE-YOLO: High Precision Recognition Algorithm for Small Target Foreign Object Detection

Illicit item detection in X-ray images for security applications

EM-YOLO: An X-ray Prohibited-Item-Detection Method Based on Edge and Material Information Fusion

A Lightweight underwater detector enhanced by Attention mechanism, GSConv and WIoU on YOLOv8

Over-sampling De-occlusion Attention Network for Prohibited Items Detection in Noisy X-ray Images

Large Span Sizes and Irregular Shapes Target Detection Methods Using Variable Convolution-Improved YOLOv8

DS-YOLOv8-Based Object Detection Method for Remote Sensing Images

Advancing Ton-Bag Detection in Seaport Logistics with an Enhanced YOLOv8 Algorithm

DBS-YOLO: A High-Precision Object Detection Algorithm for Hazardous Waste Images

SMR–YOLO: Multi-Scale Detection of Concealed Suspicious Objects in Terahertz Images