Developing Image Enhancement Algorithm for Detection of Dangerous Goods in Airport Security Inspection

Abstract

Ensuring the safety of air travel is a paramount concern, and one critical aspect is the accurate and efficient detection of dangerous goods during airport security inspections. This research presents a comprehensive approach to enhancing image data for the detection of hazardous items using advanced algorithms. The proposed methodology encompasses several key stages, including input image processing, image enhancement, and object localization employing the Shift Localization algorithm, Region of Interest (ROI)-based segmentation, and feature extraction utilizing the Gray-Level Co-occurrence Matrix (GLCM). The process begins with the pre-processing of input images to improve overall quality and prepare them for subsequent analysis. Image enhancement techniques are then applied to highlight crucial features and details relevant to the identification of dangerous goods. The Shift Localization algorithm is employed for precise object localization within the images, facilitating accurate segmentation of regions of interest. The next step involves ROI-based segmentation to isolate potential hazardous objects, followed by feature extraction using GLCM to capture textural information critical for discrimination. The extracted features are then subjected to a feature matching process to identify patterns associated with dangerous goods. To further enhance classification accuracy, a Recurrent Neural Network (RNN) classifier is employed, leveraging the temporal dependencies present in the extracted features. The proposed algorithm's performance is evaluated using standard metrics, with a primary focus on accuracy. The developed system aims to provide a robust and reliable solution for the detection of dangerous goods in airport security inspections, contributing to the overall improvement of aviation safety. The results obtained demonstrate the effectiveness of the proposed methodology, showcasing its potential for real-world application in enhancing security measures within airport environments.