Analyzing human behavior, through cameras and other monitoring technology, has been a priority for years as we look to increase our security, safety, and efficiency. Now, new innovations are arising that rethink how imaging can be facilitated and are poised to completely transform the space. Wi-Fi Doppler Imaging has emerged as a high-resolution imaging technology that leverages the Doppler Effect and standard Wi-Fi signals to identify and characterize the movements of people, pets, and objects in the home, offices and buildings.
This technology eliminates the need for multiple radar sensors for imaging and opens up new possibilities. Wi-Fi Doppler Imaging can accurately determine an object’s location and capture micro-Doppler signatures with a single Wi-Fi device to perform object labeling and classify human movement.
In this article, we’ll examine how the Doppler Effect and Doppler Spectrograms can be used in imaging and the applications this new technology has for analyzing human behavior.
What Is the Doppler Effect?
When people hear the term Doppler Effect, one of the first things that come to mind is the sound of a police car’s siren growing higher in pitch as it moves closer to an observer on the highway and lower in pitch as it moves away. This effect refers to the change in frequency of a wave during the relative motion between the source and the observer.
This frequency shift as reflected by various parts of a moving object is represented in Doppler patterns. A Doppler signature coming off of a complex object consists of a combination of micro-Doppler signatures which can be used to distinguish between kids and adults and can even label different body parts. This object labeling can then be used to differentiate human movements such as walking, sitting, standing up, and falling down.
What Is a Doppler Spectrogram?
While the Doppler Effect refers to a single frequency change, a Doppler Spectrogram is a representation of the range of Doppler frequencies generated by a moving object over time. Different types of movements produce different spectrograms.
Figure 1. A spectrogram of a person walking and then slipping backward
Spectrograms are fundamental to the analysis of movement, as well as the behavior of objects. Fields such as seismology, medicine, and music use spectrograms extensively. Like the spectrogram above, Doppler signal analytics don’t capture the actual image of a person. It only uses classifiers to identify a type of movement and create a two-dimensional illustration of the person’s movement . This allows spectrograms to analyze and classify a wide variety of movements.
What Types of Movement Can Spectrograms Capture?
Each part of the human body has characteristic movements that produce unique Doppler signatures. Figure 2 depicts a two-dimensional illustration and the three-dimensional body it represents.
Figure 2. A spectrogram illustrating a full human body in walking motion
The red zigzag pattern depicted in the spectrogram shows that the human torso’s velocity isn’t constant; it accelerates and decelerates as the person walks. It’s also important to note that the zigzag pattern isn’t unique to this particular human movement. Objects that move in a pendulum-like motion would produce such a pattern on a Doppler spectrogram.
One example of spectrograms capturing the movements of various body parts is illustrated in Figure 3, where the lower intensity spikes manifesting around the top edge of the illustration depict arm and leg movements.
Figure 3. A spectrogram of the arm movements of a person in walking motion
Looking at each contribution separately, it becomes more apparent that the high-velocity spikes are caused by arm movements. Its overall return might be weaker, but at certain times during the walking cycle, its velocity is higher than the torso’s.
Doppler Spectrograms Use in Imaging Applications
Now that the ability for Doppler Spectrograms to analyze and classify human movement has been explored, the technology can be applied to imaging applications. Existing imaging technologies have issues with field of view and the inability to see through walls and sense vital signs. Doppler Spectrograms have none of these limitations.
Combining the insights of spectrograms with machine learning algorithms , enables imaging applications to classify the contribution of each body part and detect irregularities in a person’s movement. In doing so, imaging can provide insights into human behavior, posture, and physical state.
This technology can further enrich imaging by identifying breathing patterns and sensing of various vital signs, which will be able to provide inputs into healthcare systems.
How Can Imaging Technology Analyze Human Behavior?
As mentioned, machine learning has made it possible to distinguish movement among humans, pets, or inanimate objects like a robot vacuum cleaner. When spectrogram analytics is performed over a radar on 5 or 6 GHz, it can even accurately analyze the Doppler signature of humans and identify actions such as sitting down, walking, standing, and bending over.
Figure 4. A spectrogram depicting a change in movement from walking to falling over
By leveraging Wi-Fi standard 5 GHz bands, this Imaging technology can “see” through walls. Line of sight isn’t a necessary requirement for the accurate identification and classification of human behavior within a given home environment. Thanks to Celeno’s Wi-Fi Doppler Imaging technology, even homeowners with a standard Wi-Fi router can leverage it to uncover a new range of applications and services.
Significant Applications of Wi-Fi Doppler Imaging
Combining imaging technology with Wi-Fi routers has led to a powerful innovation that can identify and classify the movement of people and objects without the aid of multiple sensors or radars for home, industrial or commercial applications.
Wi-Fi Doppler Imaging is the first of its kind. It offers an affordable and hassle-free alternative to expensive sophisticated hardware, yet it delivers more functionalities and features. Pushing the boundaries of Wi-Fi technology, Wi-Fi Doppler Imaging unlocks a new roster of applications that can improve the quality of life.
Wi-Fi Doppler Imaging can heighten home security by detecting intruders when homeowners aren’t around and even distinguish intruders from the homeowners. Since Wi-Fi Doppler Imaging technology can classify body movements, its applications can be quite helpful in home elderly care and assisted living facilities. It can notify the right people in the case that the grandmother falls or slips accidentally, allowing them to respond immediately and efficiently. Celeno’s early stage experiments, show Wi-Fi Doppler Imaging’s classification of a person falling is at least 97% accurate.
In addition to the home applications, Wi-Fi Doppler Imaging technology has applications that can empower businesses to improve healthcare, become more energy efficient, and even enhance emergency response. Companies can set up zones using geofencing. In times of an emergency, the system can identify people who have yet to exit the building and where they are.