X-Ray Image Detection Algorithm for Prohibited Items Based on Feature Enhancement and Loss Optimization

Journal Title: Engineering and Technology Journal - Year 2024, Vol 9, Issue 08

Abstract

To address the issue of low detection accuracy of prohibited items in X-ray security images caused by varying orientations, different scales, and the intertwining of targets with backgrounds, we propose a novel X-ray image detection algorithm based on feature enhancement and loss optimization. The model is built upon the ConvNext network and incorporates a Directional Channel Attention (DCA) mechanism, which efficiently captures the interaction information of local channels in different directions, thereby enhancing the accuracy of the detection model for prohibited items. Additionally, a multi-scale fusion bypass (MFB) branch is designed after the backbone network to integrate information from feature maps at different layers, thereby mitigating the interference of scale variations on the model. Furthermore, the loss function is redesigned to enable the model to automatically adjust its focus on hard samples, improving the overall detection performance. Experimental results on the SIXray dataset demonstrate that the proposed model achieves a mean Average Precision (mAP) of 91.43%, representing a 9.17% improvement over the original algorithm, thereby validating the effectiveness of the proposed method.

Authors and Affiliations

ZHAOXiaotao, LIXinwei,

Keywords

Related Articles

ANALYSIS OF MARSHALL CHARACTERISTICS WITH FLY ASH MATERIAL FOR STONE DUST SUBSTITUTION AS FILLER IN HRS-WC MIXTURE

Lataston asphalt mix is a thin layer of asphalt concrete that is often used on light-traffic roads because it produces roads with good flexibility and durability. An economical way of doing this is to vary the asphalt mi...

LOAD AND DEFLECTION ANALYSIS OF DIFFERENT BEAM

In this study, experimental analysis of various beam which were strengthened with the use of GFRP laminates was carried out and after that the result were compared with the non-strengthened beam called control beam. Thre...

A CASE STUDY OF INDUSTRIAL CONTROL SYSTEM NETWORK SECURITY

This case study presented the design of an Industrial Control System (ICS). The design focused on network security using network segmentation and redundancy. The objective was to relate theoretical network defense stra...

ASSESSMENT OF TRACE AND RARE EARTH ELEMENT DISTRIBUTION AND SPECIATION IN BAUXITE DEPOSITS ACROSS EKITI AND ONDO STATES, NIGERIA

The growing significance of rare earth elements (REEs) in advanced electronic technologies and the nuclear industry, coupled with concerns about geopolitical dynamics and supply market vulnerabilities, has reignited glob...

LEVEL OF KNOWLEDGE AND PRACTICE OF ORTHOTICS IN THE DIABETIC FOOT MANAGEMENT IN NIGERIA

Orthotists are trained to have the specialist skill and ability to make prescriptions as well as fabricate devices to achieve these offloading goals in the management of the diabetic foot. The IWGDF has been promoting so...

Download PDF file
  • EP ID EP742938
  • DOI 10.47191/etj/v9i08.28
  • Views 27
  • Downloads 0

How To Cite

ZHAOXiaotao, LIXinwei, (2024). X-Ray Image Detection Algorithm for Prohibited Items Based on Feature Enhancement and Loss Optimization. Engineering and Technology Journal, 9(08), -. https://europub.co.uk./articles/-A-742938