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.

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

Automotive radars, along with other sensors such as lidar, (which stands for "light detection and ranging"), ultrasound, and cameras, form the backbone of self-driving cars and advanced driver assistant systems (ADASs). These technological advancements are enabled by extremely complex systems with a long signal processing path from radars/sensors to the controller. Automotive radar systems are responsible for the detection of objects and obstacles, their position, and speed relative to the vehicle. The development of signal processing techniques along with progress in the millimeter-wave (mm-wave) semiconductor technology plays a key role in automotive radar systems. Various signal processing techniques have been developed to provide better resolution and estimation performance in all measurement dimensions: range, azimuth-elevation angles, and velocity of the targets surrounding the vehicles. This article summarizes various aspects of automotive radar signal processing techniques, including waveform design, possible radar architectures, estimation algorithms, implementation complexity-resolution trade off, and adaptive processing for complex environments, as well as unique problems associated with automotive radars such as pedestrian detection. We believe that this review article will combine the several contributions scattered in the literature to serve as a primary starting point to new researchers and to give a bird's-eye view to the existing research community.

Automotive Radars: A review of signal processing techniques

This paper highlights the extraction of micro-Doppler (m-D) features from radar signal returns of helicopter and human targets using the wavelet transform method incorporated with time-frequency analysis. In order for the extraction of m-D features to be realised, the time domain radar signal is decomposed into a set of components that are represented at different wavelet scales. The components are then reconstructed by applying the inverse wavelet transform. After the separation of m-D features from the target's original radar return, time-frequency analysis is then used to estimate the target's motion parameters. The autocorrelation of the time sequence data is also used to measure motion parameters such as the vibration/rotation rate. The findings show that the results have higher precision after the m-D extraction rather than before it, since only the vibrational/rotational components are employed. This proposed method of m-D extraction has been successfully applied to helicopter and human data.

Analysis of radar micro-Doppler signatures from experimental helicopter and human data

The rotation of structures in a target introduces additional frequency modulations on the returned signals and also generates sidebands about the center Doppler frequency of the target. In other words, the body image will be contaminated due to the interference from the rotating parts. In this paper, an imaging method for moving targets with rotating parts is presented. The method is simple to implement and is based on the Hough transform (HT), which is widely used in image processing. Using the standard HT and an extended HT, we put forward a separation method by detecting the straight lines and the sinusoids on the spectrogram, respectively. A computer simulation is given to illustrate the effectiveness of the proposed method.

Imaging of a Moving Target With Rotating Parts Based on the Hough Transform

The popularity of small unmanned aerial vehicles (UAVs) is increasing. Therefore, the importance of security systems able to detect and classify them is increasing as well. In this paper, we propose a new approach for UAVs classification using continuous wave radar or high pulse repetition frequency (PRF) pulse radars. We consider all steps of processing required to make a decision out of the raw radar data. Before the classification, the micro-Doppler signature is filtered and aligned to compensate the Doppler shift caused by the target's body motion. Then, classification features are extracted from the micro-Doppler signature in order to represent information about class at a lower dimension space. Eigenpairs extracted from the correlation matrix of the signature are used as informative features for classification. The proposed approach is verified on real radar measurements collected with X-band radar. Planes, quadrocopter, helicopters, and stationary rotors as well as birds are considered for classification. Moreover, a possibility of distinguishing different number of rotors is considered. The obtained results show the effectiveness of the proposed approach. It provides the capability of correct classification with a probability of around 92%. © Cambridge University Press and the European Microwave Association 2014.

Classification of small UAVs and birds by micro-Doppler signatures

: Mechanical vibrations or rotations (micro-motion dynamics) of structures on a target may introduce frequency modulation on the radar return from the target's body. The modulation due to this vibration or rotation is referred to as the micro-Doppler (m-D) phenomenon. In this report, the m-D effect is introduced and the mathematics of micro-Doppler signatures, induced by simple sinusoidal vibrations or rotations, is developed. Simulated results confirm that the mathematical analysis is valid. The m-D features derived from a target's vibrational/rotational motion are extracted by utilizing discrete wavelet transforms. During this process, the time domain radar signal is decomposed into a set of components that are represented by different wavelet scales. The m-D features are extracted by sorting the components that are associated with the vibrational/rotational motions of a target and is achieved by applying the inverse wavelet transform. After the extraction of m-D features, time-frequency analysis is employed to analyze the oscillation and to estimate the motion parameters. The vibration/rotation rate is estimated by taking the autocorrelation of the time sequence data. The findings show that these results have higher precision after the m-D extraction since only vibrational/rotational components are employed. The proposed method of the m-D extraction has been successfully applied to both simulated data and experimental helicopter and human data. The preliminary results clearly demonstrate that the m-D signatures can be observed by radar and suggest that applications of m-D should be investigated and exploited for target detection, classification and recognition. It is recommended %hat the exploitation of micro-Doppler, as a new identification/recognition tool, be undertaken as it could impact all aspects of radar sensing and may enhance the effectiveness of Automatic Target Recognition (ATR) and Automatic Gait Recognition (AGR) techniques. We recommend that

/pdf/micro-doppler-radar-signatures-for-itelligent-target-3bzmm27dc7.pdf

S. Abrol

Papers

Analysis of radar micro-Doppler signatures from experimental helicopter and human data

Micro-Doppler Radar Signatures for Itelligent Target Recognition