Radar systems are increasingly being employed in healthcare applications for human activity recognition due to their advantages in terms of privacy, contactless sensing, and insensitivity to lighting conditions. The proposed classification algorithms are however often complex, focusing on a single domain of radar, and requiring significant computational resources that prevent their deployment in embedded platforms which often have limited memory and computational resources. To address this issue, we present an adaptive magnitude thresholding approach for highlighting the region of interest in the multi-domain micro-Doppler signatures. The region of interest is beneficial to extract salient features, meanwhile it ensures the simplicity of calculations with less computational cost. The results for the proposed approach show an accuracy of up to 93.1% for six activities, outperforming state-of-the-art deep learning methods on the same dataset with an over tenfold reduction in both training time and memory footprint, and a twofold reduction in inference time compared to a series of deep learning implementations. These results can help bridge the gap toward embedded platform deployment.

Population ageing has become a severe problem worldwide. Human activity recognition (HAR) can play an important role to provide the elders with in-time healthcare. With the advantages of environmental insensitivity, contactless sensing and privacy protection, radar has been widely used for human activity detection. The micro-Doppler signatures (spectrograms) contain much information about human motion and are often applied in HAR. However, spectrograms only interpret magnitude information, resulting in suboptimal performances. We propose a radar-based HAR system using deep learning techniques. The data applied came from the open dataset “Radar signatures of human activities” created by the University of Glasgow. A new type of hybrid map was proposed, which concatenated the spectrograms amplitude and phase. After cropping the hybrid maps to focus on useful information, a convolutional neural network (CNN) based on LeNet-5 was designed for feature extraction and classification. In addition, the idea of transfer learning was applied for radar-based HAR to evaluate the classification performance of a pre-trained network. For this, GoogLeNet was taken and trained on the newly-produced hybrid maps. These initial results showed that the LeNet-5 CNN using only the spectrograms obtained an accuracy of 80.5%, while using the hybrid maps reached an accuracy of 84.3%, increasing by 3.8%. The classification result of transfer learning using GoogLeNet was 86.0%.

Radar is now widely used in human activity classification because of its contactless sensing capabilities, robustness to light conditions and privacy preservation compared to plain optical images. It has great value in elderly care, monitoring accidental falls and abnormal behaviours. Monostatic radar suffers from degradation in performance with varying aspect angles with respect to the target. Bistatic radar may offer a solution to this problem but finding the right geometry can be quite resource-intensive. We propose a bespoke simulation framework to test the radar geometry for human activity recognition. First, the analysis focuses on the monostatic radar model based on the Doppler effect in radar. We analyse the spectrogram of different motions by Short-time Fourier analysis (STFT), and then the classification data set was built for feature extraction and classification. The results show that the monostatic radar system has the highest accuracy, up to 98.17%. So, a bistatic radar model with separate transmitter and receiver was established in the experiment, and results show that bistatic radar with specific geometry configuration (CB2.5) not only has higher classification accuracy than monostatic radar in each aspect angle but also can recognise the object in a wider angle range. After training and fusing the data of all angles, it is found that the accuracy, sensitivity, and specificities of CB2.5 have 2.2%, 7.7% and 1.5% improvement compared with monostatic radar.

Radar has long been considered an important technology for indoor monitoring and assisted living. As ageing has become a worldwide problem, it causes a huge burden on the government’s healthcare expenses and infrastructure. Radar-based human activity recognition (HAR) is foreseen to become a widespread sensing modality for health monitoring at home. Conventional radar-based HAR task usually adopts the amplitude of spectrograms as input to a convolutional neural network (CNN), which can limit the achieved performances. A hybrid fusion model is here proposed, which can integrate multiple radar data domains. The result shows that the proposed framework can achieve superior classification accuracy of 92.1% (+2.5% higher than conventional CNN) and a lighter computational load than the state-of-the-art techniques with 3D-CNN.

Radar micro-Doppler signatures have been proposed for human monitoring and activity classification for surveillance and outdoor security, as well as for ambient assisted living in healthcare-related applications. A known issue is the performance reduction when the target is moving tangentially to the line of sight of the radar. Multiple techniques have been proposed to address this, such as multistatic radar and to some extent, interferometric (IF) radar. A simulator is presented to generate synthetic data representative of eight radar systems (monostatic, circular multistatic and in-line multistatic [IM] and IF) to quantify classification performances as a function of aspect angles and deployment geometries. This simulator allows an unbiased performance evaluation of different radar systems. Six human activities are considered with signatures originating from motion-captured data of 14 different subjects. The classification performances are analysed as a function of aspect angles ranging from 0° to 90° per activity and overall. It demonstrates that IF configurations are more robust than IM configurations. However, IM performs better at angles below 55° before IF configurations take over.

Radar systems are increasingly being used for healthcare applications for human activity recognition due to their advantages for privacy compliance, contactless sensing, and insensitivity to lighting conditions. The proposed classification algorithms are often very complex, hence requiring significant computational resources. We propose an adaptive thresholding algorithm used as a 'mask' to highlight the region of interest from the micro-Doppler signature. The mask is then applied to spectrogram information. These masked signatures are used for handcrafted feature extraction and classification. A quadratic-SVM classifier is employed based on the features from the information acquired. The preliminary results show that an accuracy of 91.3% is achieved using sequential forward feature selection with feature fusion. Based on our initial result, a Naïve Bayes combiner is used to improve the overall performance further. With this strategy, the accuracy of classification reaches 92.5% for six activities. Additionally, we compare our findings to those of other models utilizing the same database. The results demonstrate that high accuracy can be achieved when adaptive thresholding is used with the SVM method, and computational resources may significantly decrease.

Human Activity Classification with radar has made significant progress in the past few years. In this article, we propose a cyclostationarity-based approach in this field of application. Feature extraction, selection, and activity classification as it detects micro-Doppler is made starting from complex-valued cyclostationary statistical functions of the reflected radar signal. The human activity can be recognized with up to 92.6% with the real part, 95.4% with the imaginary part and 95.4% by the combination of real and imaginary part.

The latest progress of the multiple-input multiple-output (MIMO) radar system developed for small drones detection at Beijing Institute of Technology is presented herein. A low-cost S-band MIMO scanning radar system is designed for the detection of small drones. A practical design optimisation and implementation of a sparse array covering the whole airspace, and a target-based MIMO radar array calibration method to improve the signal-to-noise ratio of the target are proposed. An experiment in the microwave chamber was conducted to verify the designed sparse-array beam-pattern, and measurement campaigns for small drones were performed to test the system performance where the radar achieved detection and tracking of a small DJI-Phantom 4 drone at a range of up to 5 km. Target features, including micro-Doppler and radar cross-section signatures were extracted to assist target classification. The experimental results indicate that the system has promising capability in small drones detection and tracking, despite the challenges of low dwell time on targets due to the scanning nature of the system.

Radar-based classification of human activities and gait have attracted significant attention with a large number of approaches proposed in terms of features and classification algorithms. A common approach in activity classification attempts to find the algorithm (features plus classifier) that can deal with multiple activities analysed in one study such as walking, sitting, drinking and crawling. However, using the same set of features for multiple activities can be suboptimal per activity and not take into account the diversity of kinematic movements that could be captured by diverse features. In this paper, we propose a hierarchical classification approach that uses a large variety of features including but not limited to energy features like entropy and energy curve, physical features like centroid and bandwidth, image-based features like skewness extracted from multiple radar data domains. Feature selection is used at each step of the hierarchical model to select the best set of features to discriminate the target activity from the others, showing improvements with respect to the more conventional approach of using a multiclass model. The proposed approach is validated on a large dataset with 1078 recorded samples of varying length from 5 s to 10 s of experimental data, yielding 95.4% accuracy to classify six activities. The approach is also validated on a personnel recognition task to identify individual subjects from their walking gait, yielding 83.7% accuracy for ten subjects and 68.2% for a significantly larger group of subjects, i.e., 60 people.

Radar-based human activities recognition is still an open problem and is a key to detect anomalous behaviour for security and health applications. Deep learning networks such as convolutional neural networks (CNN) have been proposed for such tasks and showed better performance than traditional supervised learning paradigm. However, it is hard to deploy CNN networks to embedded systems due to the limited computational power available. From this point of concern, the use of a recurrent neural network (RNN) is proposed in this paper for human activities classification. We also propose an innovative data argumentation method to train the neural network using a limited number of data. The experiment shows that our network can achieve a mean accuracy of 94.3% in human activity classification.

Worldwide the ageing population is increasing, and there are new requirements from governments to keep people at home longer. As a consequence assisted living has been an active area of research, and radar has been identified as an emerging technology of choice for indoor activity monitoring. Activity classification has been investigated, but is often limited by the classification accuracy in the most challenging yet realistic cases. This paper aims to evaluate and improve the accuracy in classifying six commonly performed indoor activities from the University of Glasgow open dataset. For activity classification, the selection of features to discriminate between activities is paramount. Activity classification is usually done as one vs all strategy with one classifier and a set of features to distinguish between all the activities. In this paper, we propose to optimise the feature selection and classifier choice per activity using a hierarchical classification structure. This strategy reached 95.4% accuracy for all activities and about 100% for walking, opening the field for personnel recognition.

Recognition of human movements with radar for ambient activity monitoring is a developed area of research that yet presents outstanding challenges to address. In real environments, activities and movements are performed with seamless motion, with continuous transitions between activities of different duration and a large range of dynamic motions, compared with discrete activities of fixed-time lengths which are typically analysed in the literature. This paper proposes a novel approach based on recurrent LSTM and Bi-LSTM network architectures for continuous activity monitoring and classification. This approach uses radar data in the form of a continuous temporal sequence of micro-Doppler or range-time information, differently from from other conventional approaches based on convolutional networks that interpret the radar data as images. Experimental radar data involving 15 participants and different sequences of 6 actions are used to validate the proposed approach. It is demonstrated that using the Doppler-domain data together with the Bi-LSTM network and an optimal learning rate can achieve over 90% mean accuracy, whereas range-domain data only achieved approximately 76%. The details of the network architectures, insights in their behaviour as a function of key hyper-parameters such as the learning rate, and a discussion on their performance across are provided in the paper.