Human Activity Classification Based on Micro-Doppler Signatures Using a Support Vector Machine

TLDR

The study investigates the feasibility of classifying human activities from micro‑Doppler signatures and explores its potential for extended duration, through‑wall, and oblique‑angle scenarios. Data from 12 subjects performing seven activities were collected with a Doppler radar, six spectrogram features were extracted, and a support‑vector‑machine classifier—trained via four‑fold cross‑validation and organized in a decision‑tree structure—was used for multiclass activity recognition. The classifier achieved over 90 % accuracy, demonstrating the viability of micro‑Doppler based activity recognition.

Abstract

The feasibility of classifying different human activities based on micro-Doppler signatures is investigated. Measured data of 12 human subjects performing seven different activities are collected using a Doppler radar. The seven activities include running, walking, walking while holding a stick, crawling, boxing while moving forward, boxing while standing in place, and sitting still. Six features are extracted from the Doppler spectrogram. A support vector machine (SVM) is then trained using the measurement features to classify the activities. A multiclass classification is implemented using a decision-tree structure. Optimal parameters for the SVM are found through a fourfold cross-validation. The resulting classification accuracy is found to be more than 90%. The potentials of classifying human activities over extended time duration, through wall, and at oblique angles with respect to the radar are also investigated and discussed.