Extraction, Based On Human Detection And Gait Characteristics Of The Radar Echoes

Since the human beings and the other creatures usually walk or run with a constant velocity, stride and gait frequency, there are some Micro-Doppler signatures on their spectrums. Human gait detection and characteristic extraction mainly utilize the walking human's dynamic features to modulate the radar echo phase, use time-frequency analysis method to obtain the human spectrum, and extract the Micro-Doppler signature of human gait from spectrum. After detecting the human gait, extraction of gait parameters including velocity, stride, gait frequency, et al are accomplished finally.This paper proposes a new human walking Doppler model, researches the target detection, spectrum analysis, gait parameters estimation, as well as gender discrimination of the walking human. And the main problems are studied as below:1. By utilizing a 12-point model of human beings, a new signal model based on a continuous wave and Linear Frequency Modulation (LFM) pulses is proposed, which is combined with different radar scenes for walking human target.2. An estimation of walking human's range is researched. The echo pulses in each range cell pass through a Moving Target Detection (MTD) filter firstly, then a Constant False Alarm Ratio (CFAR) algorithm in frequency domain after coherent accumulation is used to detection the target based on the LFM radar signal system. The detection performance is given by analyzing the influence of the slipping window and the false alarm.3. The problem of target span range cells is analyzed, a real-time detecting method is given to detect the walking human's range and estimate its velocity. So movement compensation can be done to the echo signal.4. A time-frequency analysis method - the Short Time Fourier Transform (STFT) - is used to obtain the human spectrum. The Micro-Doppler signatures of different body parts are deduced mathematically, and the results are approved correct by comparing with the measured spectrum from some experimental data. At the same time, the new signal model of walking human is also approved correct.5. An algorithm based on Fast Fourier Transform (FFT) to extract the cadence frequency is used to estimate the human target's gait frequency. It mainly utilizes the bandwidth and the variation period of the Doppler shift to calculate the target's velocity and gait frequency.6. A gender discrimination method is researched in theory. Since the length of stride and the height of thigh can be estimated from velocity and gait frequency by an empirical formula, the target's gender can be judged by utilizing the Possibility Distribution Function (PDF) of man and woman's height of thigh.The theoretical validities of human gait detection and characteristic extraction are illustrated with simulation results. It can be widely used in life medicine, rescue service, damage detection and so on. It also can make up some deficiencies that the ordinary detection of living things can't distinguish human beings from the other creatures. And it's also useful to the further study, such as human targets classification, status classification and so on.