Coded-Aperture Computational Millimeter-Wave Image Classifier Using Convolutional Neural Network

Rahul Sharma,Mohsen Khalily,Andreas Keil,Okan Yurduseven,Vincent Fusco,Fabian Friederich,Thomas Fromenteze,Bhabesh Deka,Raphael Hussung

doi:10.1109/access.2021.3107782

Abstract

A millimeter-wave (mmW) classifier system applied to images synthesized from a coded-aperture based computational imaging (CI) radar is presented. A developed physical model of a CI system is used to generate the image dataset for the classification algorithm. A convolutional neural network (CNN) is integrated with the physical model and trained using the dataset comprising of synthesized mmW images obtained directly from the developed CI physical model. A ${k}$ -fold cross validation technique is applied during the training process to validate the classification model. The coded-aperture CI concept enables image reconstruction from a significantly reduced number of back-scattered measurements by facilitating physical layer compression. This physical layer compression can substantially simplify the data acquisition layer of imaging radars, which is realized using only two channels in this article. The integration of the classification algorithm with the CI numerical model is particularly important in enabling the training step to be carried out using relevant system metrics and without the necessity for experimental data. Leveraging the CI numerical model generated data, training step for the classification algorithm is achieved in real-time while also confirming that the numerically trained CI classifier offers high accuracy with both simulated and experimental data. The classifier integrated physical model also enables performance analysis of the classification algorithm to be carried out as a function of key system metrics such as signal-to-noise (SNR) level, ensuring a complete understanding of the classification accuracy under different operating conditions. The trained CI system is tested with synthesized mmW images from the physical model and a classification accuracy of 89% is achieved. The proposed model is also verified using experimental data validating the fidelity of the developed CI integrated classifier system. A classification latency of 3.8 ms per frame is achieved, paving the way for real-time automated threat detection (ATD) for security-screening applications.

Highlights

M ILLIMETER-WAVE imaging has found applications in various areas ranging from remote sensing [1]–[3] to autonomous robotics [4]–[6] and medical applications such as breast cancer detection [7], [8]
This paper shows a convolutional neural network (CNN)-enabled classification algorithm integrated into a coded-aperture computational imaging (CI) model
Our system has an advantage over such approaches in the sense that we propose an end-to-end mmW CI model for automated threat detection (ATD) in real-time

Summary

INTRODUCTION

M ILLIMETER-WAVE (mmW) imaging has found applications in various areas ranging from remote sensing [1]–[3] to autonomous robotics [4]–[6] and medical applications such as breast cancer detection [7], [8]. The developed CI model can significantly simplify the hardware layer of conventional radar architectures due to the single-frequency physical layer compression (Section IV), it paves the way for physical model development for imaging systems with deep learning to facilitate real-time classification (Sections V and VI) This is important for security-screening applications, as real-time classification is the key element for automated threat detection (ATD). The main motivation of this work is to develop a codedaperture based mmW imaging system that facilitates compressive sensing by leveraging physical layer compression and its integration with a CNN-based classifier as an enabling technology for ATD in real-time.

RELATED WORK

CNN CLASSIFIER

CNN TRAINING

CLASSIFICATION RESULTS

PRESENCE OF NOISE AND CLASSIFICATION

VIII. CONCLUSION