CN105629228A - Partition human body motion detection method based on k-means clustering and Bayes classification - Google Patents

Partition human body motion detection method based on k-means clustering and Bayes classification Download PDF

Info

Publication number
CN105629228A
CN105629228A CN201610042027.6A CN201610042027A CN105629228A CN 105629228 A CN105629228 A CN 105629228A CN 201610042027 A CN201610042027 A CN 201610042027A CN 105629228 A CN105629228 A CN 105629228A
Authority
CN
China
Prior art keywords
signal
range
partition wall
difference value
extreme difference
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN201610042027.6A
Other languages
Chinese (zh)
Other versions
CN105629228B (en
Inventor
张志浩
史治国
陈积明
程鹏
王�琦
孙优贤
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Zhejiang University ZJU
Original Assignee
Zhejiang University ZJU
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Zhejiang University ZJU filed Critical Zhejiang University ZJU
Priority to CN201610042027.6A priority Critical patent/CN105629228B/en
Publication of CN105629228A publication Critical patent/CN105629228A/en
Application granted granted Critical
Publication of CN105629228B publication Critical patent/CN105629228B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
    • G01S13/88—Radar or analogous systems specially adapted for specific applications
    • G01S13/887—Radar or analogous systems specially adapted for specific applications for detection of concealed objects, e.g. contraband or weapons
    • G01S13/888—Radar or analogous systems specially adapted for specific applications for detection of concealed objects, e.g. contraband or weapons through wall detection

Landscapes

  • Engineering & Computer Science (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Physics & Mathematics (AREA)
  • Electromagnetism (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • General Physics & Mathematics (AREA)
  • Image Analysis (AREA)
  • Radar Systems Or Details Thereof (AREA)

Abstract

The invention discloses a partition human body motion detection method based on k-means clustering and bayesian classification. The partition human body motion detection method is characterized in that a receiver is used to receive a signal waveform after precoding of a transmitter; the received signal can be segmented into signal segments, and the short-time Fourier transformation of each signal segment can be carried out to acquire a transformation matrix, and the variance vector of the transformation matrix can be calculated to acquire the range; and the range can be disposed in a Bayes classifier for the classification, and the Bayes classifier is acquired by adopting the clustering of the training data in advance and the Bayes classification; according to the classification result, the motion of the partition human body can be determined. By adopting the k-means clustering and the Bayes classification, whether the partition human body moves can be effectively detected, and therefore the accuracy of the partition human body motion detection can be greatly improved.

Description

Partition wall body movement detection method based on K mean cluster and Bayes's classification
Technical field
The present invention relates to a kind of partition wall body movement detection method, more specifically a kind of partition wall body movement detection method based on K mean cluster and Bayes's classification.
Technical background
Human detection in general sighting distance, it is possible to use the optoelectronic devices such as such as infrared, video camera detect. These technology are common in the intrusion detection of odeum and bank. But these technology have significant limitation, it is impossible to the detection of non-transparent medium body of wall (or veil) rear objects such as competent, concrete wooden for stone, so the detection technique adopted need to have transparent effect. Have that the detection technique of transparent effect is common to be had based on the mode such as X ray and ultrasonic echo at present, but these several fluoroscopy techniques all can not be well adapted for demand for human detection through walls at present. X ray belongs to high-energy rays, although can penetrate body of wall, but human body has very big injury; And the medium of layering is had relatively larger decay by ultrasonic echo. In sum, adopt body of wall is had good penetration, transmitting signal that human injury's negligible characteristic frequency electromagnetic wave is detected as partition wall human motion has good feasibility. Electromagnetic wave, as launching signal, can penetrate the nonmetal medium such as timber, concrete walls, it is achieved to the detection of moving target after wall.
In the particular action such as anti-riot and emergency relief, can effectively detect the body motion information or after wall will on the impact fought and rescue generation will be great in the room in, it is possible to reduce the number of casualties significantly. Therefore, it is possible to wall, timber etc. is nonmetal, transparent medium rear object detection technique is of increased attention.
Although traditional ULTRA-WIDEBAND RADAR through walls is capable of the detection of partition wall human motion, but it takies substantial amounts of bandwidth, and transmitting power is big, and has very big aerial array. And occupied bandwidth is little, transmitting power is low, the Wireless Telecom Equipment of small volume realizes partition wall human motion detection and has very big challenge, will realize the detection of weak signal target under very noisy. The technology of the partition wall body movement detection method realized about this portable set at present needs to be studied and discussion.
Summary of the invention
It is an object of the invention to propose a kind of partition wall body movement detection method based on K mean cluster and Bayes's classification, it is possible to effectively detect whether partition wall human body moves, improve detection accuracy.
It is an object of the invention to be achieved through the following technical solutions: a kind of partition wall body movement detection method based on K mean cluster and Bayes's classification, the method comprises the following steps:
Step 1, arranges the first transmitter, the second transmitter and receiver in the side of wall; First the first transmitter sends primary signal, and after receiver receives signal, the second transmitter sends same primary signal, and receiver receives signal; The signal then passing through twice reception calculates the precoded signal of the second transmitter; Signal launched by last two transmitters simultaneously, and the first transmitter sends primary signal, and the second transmitter sends precoded signal;
Step 2, receiver receives the signal after the superposition that two transmitters send simultaneously, and temporally carries out even partition to the received signal;
Step 3, carries out Short Time Fourier Transform to every segment signal of step 2 segmentation, obtains a Short Time Fourier Transform matrix Am��n, m represents the Frequency point number of Fourier transformation (FFT), and n is the time point number according to window function size and the overlapping calculated every segment signal of number, the elements A in matrixijRepresent in i frequency, the Short Time Fourier Transform value of j time point;
Step 4, the Short Time Fourier Transform matrix A that step 3 is obtainedm��nCarry out variance statistic, namely calculate the variance v of Short Time Fourier Transform value corresponding to all Frequency points on each time pointj, finally give the variance vectors v on all time points of this segment signal1��n; Simultaneously to Short Time Fourier Transform matrix Am��nCarry out median absolute deviation statistics, obtain median absolute deviation vector M AD1��n;
Step 5, calculates variance vectors v1��nExtreme difference value vrange, i.e. vrange=vmax-vmin, vmaxFor variance vectors v1��nIn maximum, vminFor variance vectors v1��nIn minima; In like manner calculate median absolute deviation vector M AD1��nExtreme difference value MADrange;
Step 6, calculates the extreme difference value v ' when partition wall has people to move according to step 1-5 respectivelyrange��MAD'rangeExtreme difference value v during motion unmanned with partition wall "range��MAD"range; Adopt K average (Kmeans) method that the extreme difference value in two kinds of situations is carried out cluster and aggregate into two bunches, and extreme difference value and cluster result are carried out Bayes's classification as training set, obtain a Bayes classifier;
Step 7, when carrying out partition wall human motion detection, calculates the extreme difference value v of a segment signal according to step 1-5rangeAnd MADrange, by extreme difference value vrangeAnd MADrangePutting into the Bayes classifier that step 6 obtains to classify, if Bayes classifier is categorized into partition wall human motion situation, then this moment partition wall human body is in motion; And being classified as another kind of, then partition wall does not have human body in motion; Every segment signal of step 2 segmentation is repeated this step, such that it is able to provide the moment of partition wall human motion.
Partition wall body movement detection method based on K mean cluster and Bayes's classification of the present invention, it is possible to detect whether partition wall has human motion. Compared with prior art, the present invention has the advantage that
1, adopt the similarity of K mean cluster and Bayes's classification signal calculated waveform, compare threshold test and other detection methods, it is not necessary to go to select suitable threshold value; But classified by grader, the standard of classification is determined by grader;
2, real-time detection can be realized, carry out corresponding signal processing according to the signal received, and provide in real time the motion that detects towards result;
3, different environment and different human motion patterns it are adapted to, without changing accordingly for the change of environment and motor pattern in advance;
4, check frequency is little, can realize detection in effective detection region.
Accompanying drawing explanation
Fig. 1 is the flow chart of transmitter and receiver;
Fig. 2 is based on the partition wall human motion detection signal processing flow figure of K mean cluster and Bayes's classification;
Fig. 3 is K mean cluster result.
Detailed description of the invention
Below in conjunction with accompanying drawing, the present invention is described in further detail.
The present invention gives a kind of partition wall body movement detection method based on K mean cluster and Bayes's classification, transmission and the reception process of signal are as it is shown in figure 1, used is two transmitters and a receiver. First, the first transmitter sends signal, and receiver receives signal; Secondly the second transmitter sends the signal same with the first transmitter, and receiver receives signal; Then according to the signal received for twice, the signal after precoding is calculated; Most two transmitters of relief send signal simultaneously, and receiver receives signal. Here the first transmitter still sends original signal, and the second transmitter is then the signal after sending the precoding just calculated.
On the basis that above-mentioned signal sends and receives, detection method of the present invention, as in figure 2 it is shown, comprise the following steps:
Step 1, first allows the side that receiver and two transmitters are placed on wall run a period of time, and receiver will receive after wall and the signal of wall multiple reflected signal superposition here;
Step 2, temporally carries out even partition to the received signal, is divided into a section small-signal, is specifically divided into the signal data of 1s here;
Step 3, carries out Short Time Fourier Transform (STFT) STFT (t, ��)=�� s (t') �� (t'-t) e to every section of small-signal after segmentation-j��t'Dt', obtains a Short Time Fourier Transform matrix Am��n, the line number m of this matrix represents the Fourier transformation (FFT) using how many points, namely has how many Frequency points; Matrix column number n is then the time point number according to window function size and the overlapping calculated every section of small-signal of number. Elements A so this transformation matrix is not only relevant with frequency, and also relevant with the time, in matrixijRepresent in i frequency, the Short Time Fourier Transform value of j time point;
Step 4, the Short Time Fourier Transform matrix A that step 3 is obtainedm��nCarry out variance statistic, namely calculate the variance v of Short Time Fourier Transform value corresponding to all Frequency points on each time pointj, finally give the variance vectors v on all time points of this segment signal1��n; Simultaneously to transformation matrix Am��nCarry out median absolute deviation statistics, obtain median absolute deviation vector M AD1��n;
Step 5, calculates variance vectors v1��nExtreme difference value vrange, i.e. vrange=vmax-vmin, vmaxFor variance vectors v1��nIn maximum, vminFor variance vectors v1��nIn minima; In like manner calculate median absolute deviation vector M AD1��nExtreme difference value MADrange;
Step 6, calculates the extreme difference value v ' when partition wall has people to move according to step 1-5 respectivelyrange��MAD'rangeExtreme difference value v during motion unmanned with partition wall "range��MAD"range; Adopt K average (Kmeans) method that the extreme difference value in two kinds of situations is carried out cluster and aggregate into two bunches, and extreme difference value and cluster result are carried out Bayes's classification as training set, obtain a Bayes classifier;
Step 7, when carrying out partition wall human motion detection, calculates the extreme difference value v of a segment signal according to step 1-5rangeAnd MADrange, by extreme difference value vrangeAnd MADrangePutting into the Bayes classifier that step 6 obtains to classify, if Bayes classifier is categorized into partition wall human motion situation, then this moment partition wall human body is in motion; And being classified as another kind of, then partition wall does not have human body in motion; Every segment signal of step 2 segmentation is repeated this step, such that it is able to provide the moment of partition wall human motion.
Detection method of the present invention adopts the technical scheme that: in advance training data is carried out K mean cluster and becomes two bunches, then training data and cluster result are obtained a Bayes classifier by Bayes's classification, secondly the signal received is divided into a section signal, every segment signal is carried out Short Time Fourier Transform and obtains transformation matrix, and calculate the variance vectors of transformation matrix and then obtain extreme difference value; Finally extreme difference value is put into Bayes classifier and obtain Decision Classfication result.
The present invention adopts the transmitter that bandwidth is little, transmitting power is low can realize partition wall human motion, and can ensure accuracy of detection. Taking substantial amounts of bandwidth, high emission power and very big aerial array like that compared to traditional ULTRA-WIDEBAND RADAR through walls, the present invention has significant advantage.
Embodiment
Two transmitters and receiver are arranged in the side of wall, and movement human is optionally walked at the opposite side of wall. Two transmitter and receiver equidistant arrangement in same level, and with metope apart from equal. The body of wall of experiment is the concrete walls that 25cm is thick, and it decays to 20dB. The bandwidth of transmitter is 1MHz, and transmitting power is 100mW, and tranmitting frequency is 2.4GHz, comprises 3 beam antennas. In order to make motor pattern simpler and regular, define two kinds of motor patterns, 1) it is parallel to metope walking and 2) vertical wall walking. K mean cluster result is as shown in Figure 3.
According to the inventive method, to the verification and measurement ratio of partition wall human motion up to 90%, taking substantial amounts of bandwidth, high emission power relative to traditional ULTRA-WIDEBAND RADAR through walls, the inventive method also has higher accuracy of detection when narrow bandwidth and low transmitting power.

Claims (1)

1. the partition wall body movement detection method based on K mean cluster and Bayes's classification, it is characterised in that the method comprises the following steps:
Step 1, arranges the first transmitter, the second transmitter and receiver in the side of wall; First the first transmitter sends primary signal, and after receiver receives signal, the second transmitter sends same primary signal, and receiver receives signal; The signal then passing through twice reception calculates the precoded signal of the second transmitter; Signal launched by last two transmitters simultaneously, and the first transmitter sends primary signal, and the second transmitter sends precoded signal;
Step 2, receiver receives the signal after the superposition that two transmitters send simultaneously, and temporally carries out even partition to the received signal;
Step 3, carries out Short Time Fourier Transform to every segment signal of step 2 segmentation, obtains a Short Time Fourier Transform matrix Am��n, m represents the Frequency point number of Fourier transformation (FFT), and n is the time point number according to window function size and the overlapping calculated every segment signal of number, the elements A in matrixijRepresent in i frequency, the Short Time Fourier Transform value of j time point;
Step 4, the Short Time Fourier Transform matrix A that step 3 is obtainedm��nCarry out variance statistic, namely calculate the variance v of Short Time Fourier Transform value corresponding to all Frequency points on each time pointj, finally give the variance vectors v on all time points of this segment signal1��n; Simultaneously to Short Time Fourier Transform matrix Am��nCarry out median absolute deviation statistics, obtain median absolute deviation vector M AD1��n;
Step 5, calculates variance vectors v1��nExtreme difference value vrange, i.e. vrange=vmax-vmin, vmaxFor variance vectors v1��nIn maximum, vminFor variance vectors v1��nIn minima; In like manner calculate median absolute deviation vector M AD1��nExtreme difference value MADrange;
Step 6, calculates the extreme difference value v ' when partition wall has people to move according to step 1-5 respectivelyrange��MAD'rangeExtreme difference value v during motion unmanned with partition wall "range��MAD"range; Adopt K average (Kmeans) method that the extreme difference value in two kinds of situations is carried out cluster and aggregate into two bunches, and extreme difference value and cluster result are carried out Bayes's classification as training set, obtain a Bayes classifier;
Step 7, when carrying out partition wall human motion detection, calculates the extreme difference value v of a segment signal according to step 1-5rangeAnd MADrange, by extreme difference value vrangeAnd MADrangePutting into the Bayes classifier that step 6 obtains to classify, if Bayes classifier is categorized into partition wall human motion situation, then this moment partition wall human body is in motion; And being classified as another kind of, then partition wall does not have human body in motion; Every segment signal of step 2 segmentation is repeated this step, such that it is able to provide the moment of partition wall human motion.
CN201610042027.6A 2016-01-21 2016-01-21 Partition wall body movement detection method based on K mean cluster and Bayes's classification Active CN105629228B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201610042027.6A CN105629228B (en) 2016-01-21 2016-01-21 Partition wall body movement detection method based on K mean cluster and Bayes's classification

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201610042027.6A CN105629228B (en) 2016-01-21 2016-01-21 Partition wall body movement detection method based on K mean cluster and Bayes's classification

Publications (2)

Publication Number Publication Date
CN105629228A true CN105629228A (en) 2016-06-01
CN105629228B CN105629228B (en) 2018-04-27

Family

ID=56044344

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201610042027.6A Active CN105629228B (en) 2016-01-21 2016-01-21 Partition wall body movement detection method based on K mean cluster and Bayes's classification

Country Status (1)

Country Link
CN (1) CN105629228B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP3699631A1 (en) * 2019-02-19 2020-08-26 Fujitsu Limited Living object detection method and apparatus and electronic device

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5457394A (en) * 1993-04-12 1995-10-10 The Regents Of The University Of California Impulse radar studfinder
CN202583456U (en) * 2012-04-28 2012-12-05 电子科技大学 Building perspective detection device based on hybrid waveforms
CN104820246A (en) * 2015-04-24 2015-08-05 芜湖航飞科技股份有限公司 Through-the-wall radar human body detecting device
CN105137423A (en) * 2015-09-30 2015-12-09 武汉大学 Real-time detection and separation method of multiple moving objects by through-the-wall radar

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5457394A (en) * 1993-04-12 1995-10-10 The Regents Of The University Of California Impulse radar studfinder
US5512834A (en) * 1993-05-07 1996-04-30 The Regents Of The University Of California Homodyne impulse radar hidden object locator
CN202583456U (en) * 2012-04-28 2012-12-05 电子科技大学 Building perspective detection device based on hybrid waveforms
CN104820246A (en) * 2015-04-24 2015-08-05 芜湖航飞科技股份有限公司 Through-the-wall radar human body detecting device
CN105137423A (en) * 2015-09-30 2015-12-09 武汉大学 Real-time detection and separation method of multiple moving objects by through-the-wall radar

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
CH SENG等: ""Image segmentation for through-the-wall radar target detection"", 《IEEE TRANSACTIONS ON AEROSPACE AND ELECTRONIC SYSTEMS》 *
JING LI等: ""Through-wall detection of human being’s movement by UWB radar"", 《IEEE GEOSCIENCE AND REMOTE SENSING LETTERS》 *
S S RAM等: ""Through-wall tracking of human movers using joint doppler and array processing"", 《IEEE GEOSCIENCE AND REMOTE SENSING LETTERS》 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP3699631A1 (en) * 2019-02-19 2020-08-26 Fujitsu Limited Living object detection method and apparatus and electronic device
US11378648B2 (en) 2019-02-19 2022-07-05 Fujitsu Limited Living object detection method and apparatus and electronic device

Also Published As

Publication number Publication date
CN105629228B (en) 2018-04-27

Similar Documents

Publication Publication Date Title
Kim et al. Human detection using Doppler radar based on physical characteristics of targets
US8717223B2 (en) Classification of subsurface objects using singular values derived from signal frames
CN107861123B (en) Method for real-time tracking of multiple moving targets by through-wall radar in complex environment
US20180313950A1 (en) CNN-Based Remote Locating and Tracking of Individuals Through Walls
Nakamura et al. Wi-fi-CSI-based fall detection by spectrogram analysis with CNN
CN107462873A (en) A kind of radar chaff method for quickly identifying
US20230045798A1 (en) Method for predicting structure of indoor space using radio propagation channel analysis through deep learning
CN109031277A (en) A kind of through-wall radar multi-Target Image domain robust tracking method
CN112859063A (en) Multi-human-body target recognition and counting method based on millimeter waves
Comiter et al. Localization convolutional neural networks using angle of arrival images
EP3417311B1 (en) A method for motion classification using a pulsed radar system
Sinitsyn et al. Determination of aircraft current location on the basis of its acoustic noise
CN105708471B (en) Partition wall body movement detection method based on Short Time Fourier Transform
CN107219522B (en) Ellipse-hyperbola combined through-wall radar target positioning method
CN105629228A (en) Partition human body motion detection method based on k-means clustering and Bayes classification
Iqbal et al. Indoor motion classification using passive RF sensing incorporating deep learning
CN105629227B (en) Partition wall body movement detection method based on continuous wavelet transform
WO2017188905A1 (en) A method for motion classification using a pulsed radar system
Lee et al. -10CNN-Based UAV Detection and Classification Using Sensor Fusion
Al-Hameed et al. Artificial neural network for LiDAL systems
Damyanov et al. Super-resolution feature extraction imaging algorithm for complex objects
Liang et al. An improved sensing method using radio frequency detection
CN112801065B (en) Space-time multi-feature information-based passive sonar target detection method and device
CN105738862B (en) Partition wall human motion based on dynamic time warping is towards detection method
Liu et al. UWB vital sign signal recognition method based on SVM

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant