KR20210043448A - Apparatus and method for motion detecting - Google Patents

Abstract

The present application relates to a device and method for motion detection. The device for motion detection according to one embodiment of the present invention may comprise: a sound wave generator that periodically outputs sound waves in a secure space; a sound wave receiver that measures the reflected sound waves in the secure space, and generates a plurality of sound wave signals for each frequency; an analysis unit that generates analysis data by signal processing the sound wave signals, and generates an adaptive threshold value reflecting the environmental characteristics of the secure space; and a determination unit that compares the analysis data with an adaptive threshold value to detect an operation in the secure space. Therefore, the present invention is capable of allowing for the possibility of detecting actions such as intrusions in the secure space.

Description

Motion detection device and motion detection method {Apparatus and method for motion detecting}

The present application relates to a motion detection device and a motion detection method, and to a motion detection device and a motion detection method capable of detecting a motion such as intrusion in a secure space using sound waves reflected in a secure space.

Commonly used security systems include an IR sensor (infrared sensor), an image sensor, and a heat sensor. However, in the case of an IR sensor, it is possible to become useless due to exposure of the boundary area as it cannot detect an intruder when intruding wearing an IR shielding suit. In the case of an image sensor, misinformation due to light and shadow occurs frequently, and at night Lighting equipment is required, and the heat ray sensor reacts sensitively to changes in the surrounding temperature, and there is a high probability of misinformation, so there is a blind spot as a security system.

In addition, in order to build a security system, not only the purchase cost of various sensors mentioned above, but also the cost including time and labor cost for the construction are expended, there is a problem that the general public cannot easily access it.

Therefore, when building and using a security system, time and cost are saved, no misinformation occurs, no additional device is required, and the necessity of developing a security system that can detect intruders in any form arises.

The present application is to provide a motion detection device and a motion detection method capable of detecting motions such as intrusion in a secure space by using sound waves reflected in a secure space.

An object of the present application is to provide a motion detection device and a motion detection method capable of easily distinguishing a pattern change due to environmental changes such as air flow and a pattern change due to intrusion using cosine similarity between sound wave signals.

The present application is to provide a motion detection device and a motion detection method capable of adaptively applying a threshold for detecting motion in a secure space.

A motion detection device according to an embodiment of the present invention includes: a sound wave generator for periodically outputting sound waves in a secure space; A sound wave receiver configured to measure sound waves reflected in the secure space and generate a plurality of sound wave signals for each frequency; An analysis unit for generating analysis data by signal processing the sound wave signals, and generating an adaptive threshold reflecting the environmental characteristics of the secure space; And a discriminating unit that compares the analysis data with an adaptive threshold value and detects a motion in the secure space.

Here, the sound wave generator may output a short wave signal in the secure space and then use the measured reverberation signal to set the waveform of the sound wave.

Here, the sound wave generator may reset the waveform of the sound wave of the next period by using the reverberation signal of the sound wave output in the secure space.

Here, the sound wave generator may rearrange the waveform of the reverberation signal in the reverse order of time and set it as the waveform of the sound wave.

Here, the sound wave receiver may perform Fast Fourier Transform (FFT) to generate a plurality of sound wave signals for each frequency of each sound wave.

Here, the motion detection device according to an embodiment of the present invention further includes an initial setting unit for analyzing the sound wave received by the sound wave receiving unit and adjusting the volume of the sound wave output by the sound wave generating unit or the reception sensitivity of the sound wave receiving unit can do.

Here, the analysis unit may generate the analysis data by calculating cosine similarity between the sound wave signals, respectively.

Here, the adaptive threshold may be set by using an environmental adaptation based on past data, which is analysis data of sound wave signals received in the past, and a sensitivity for adjusting a fluctuation range of the environmental adaptation. have.

Here, the analysis unit may calculate the environmental adaptation degree by collecting past data up to a preset past time point, applying a weight to each of the past data, and adding them.

Here, the analysis unit may set a range in which a difference between the environmental adaptation and the analysis data is allowed as the sensitivity.

Here, the determination unit may determine one of an event occurrence state in which a preset motion is detected in the secure space and a normal state in which no event occurs in the secure space.

Here, when the determination unit is determined to be in the normal state, the sound wave signal or the analysis data may be compared with a preset exception situation to determine whether it corresponds to an environment change state corresponding to a temperature change or an abnormal air current occurrence in the secure space.

Here, the determination unit may classify each type of environment change corresponding to the environment change state by using the sound wave signal or analysis data.

Here, when the analysis unit is determined to be the normal state, the analysis data corresponding to the normal state is added to the past data, and the adaptive threshold is based on the past data, which is analysis data of sound wave signals received in the past. It can be set using an environmental adaptation and a sensitivity that adjusts the fluctuation range of the environmental adaptation.

A motion detection method according to an embodiment of the present invention relates to a motion detection method in a secure space using a motion detection device, comprising: periodically outputting sound waves in the secure space; Measuring sound waves reflected in the secure space and generating a plurality of sound wave signals for each frequency; Generating analysis data by performing signal processing between the sound wave signals; Generating an adaptive threshold reflecting the environmental characteristics of the secure space; And comparing the analysis data with an adaptive threshold to detect motion in the secure space.

In addition, the solution to the above-described problem does not enumerate all the features of the present invention. Various features of the present invention and advantages and effects thereof may be understood in more detail with reference to the following specific embodiments.

According to the motion detection apparatus and the motion detection method according to an embodiment of the present invention, it is possible to detect an intrusion in the secure space using sound waves reflected in the secure space.

According to the motion detection device and the motion detection method according to an embodiment of the present invention, since it is possible to distinguish a pattern change of sound waves due to environmental changes such as air flow in a secure space and a pattern change due to intrusion, etc. It can reduce detection. In addition, it is possible to classify environmental changes in the secure space for each type.

1 is a block diagram showing a motion detection apparatus according to an embodiment of the present invention.
2 is a schematic diagram showing an intrusion into a secure space and detection of an environment change using a motion detection device according to an embodiment of the present invention.
3 is a graph showing a short wave signal and a reverberation signal according to an embodiment of the present invention.
4 is a graph showing a reverberation signal and a sound wave according to an embodiment of the present invention.
5 to 8 are graphs showing the performance of the motion detection device according to each embodiment.
9 and 10 are flow charts showing a motion detection method according to an embodiment of the present invention.

Hereinafter, preferred embodiments will be described in detail with reference to the accompanying drawings so that those of ordinary skill in the art may easily implement the present invention. However, in describing a preferred embodiment of the present invention in detail, if it is determined that a detailed description of a related known function or configuration may unnecessarily obscure the subject matter of the present invention, the detailed description thereof will be omitted. In addition, the same reference numerals are used throughout the drawings for parts having similar functions and functions.

In addition, throughout the specification, when a part is said to be'connected' with another part, it is not only'directly connected', but also'indirectly connected' with another element in the middle. Includes. In addition, "including" a certain component means that other components may be further included rather than excluding other components unless otherwise stated. In addition, terms such as "~ unit" and "module" described in the specification mean a unit that processes at least one function or operation, which may be implemented as hardware or software, or as a combination of hardware and software.

1 is a block diagram showing a motion detection apparatus according to an embodiment of the present invention.

Referring to FIG. 1, a motion detection device 100 according to an embodiment of the present invention includes a sound wave generator 110, a sound wave receiver 120, an initial setting unit 130, an analysis unit 140, and a determination unit ( 150) may be included.

Hereinafter, a motion detection device according to an embodiment of the present invention will be described with reference to FIG.

The sound wave generator 110 may periodically output sound waves in the secure space. As shown in Fig. 2, the sound waves propagated in the security space (s) can be reflected by structures included in the security space (s), and then, using the reflected sound waves, intrusion into the security space (s), etc. Can detect the motion of That is, the sound wave generator 110 may output sound waves in the security space (s) to detect intrusion in the security space (s), and according to the embodiment, the sound waves are periodically output to the security space ( The motion detection for s) can be repeatedly performed periodically.

Meanwhile, the sound wave generator 110 may output sound waves having various frequencies in the secure space s, and according to embodiments, sound waves having a frequency of an inaudible frequency band may be utilized. In the case of using a frequency in an inaudible frequency band, since users cannot recognize sound waves for motion detection, it is possible to perform motion detection without inconvenience even when users are located in the secure space (s). For example, the motion detection device 100 may be included in an electronic device such as a smart speaker, and at this time, sound waves in an inaudible frequency band may be utilized. In this case, the user can input a voice command to the smart speaker or receive a function such as sound or voice guidance provided by the smart speaker, and at the same time, the motion detection device 100 detects the user's movement in the secure space (s). You can receive services such as.

Depending on the embodiment, the sound wave output from the sound wave generator 110 may be a multitone sound file composed of a linear sum of sine waves having a plurality of frequency components. In this case, the sound wave generator 110 may set the size of each frequency component, and the like, so that the output sound wave may have a specific pattern in the frequency domain. For example, when using the inaudible frequency band, it is possible to form sound waves to have a specific pattern capable of performing optimal motion recognition in the inaudible band.

In addition, when the sound wave generator 110 outputs sound waves in the secure space (s), the reflected sound waves may be affected by a temperature change or air flow inside the secure space (s). For example, as shown in FIG. 2, when the air conditioner in the secure space (s) operates, the transmission speed of sound waves and the like may change according to the internal temperature. That is, the higher the temperature, the higher the transmission speed, so that the frequency of the sound wave can shift in the high-frequency direction, and the lower the temperature, the lower the transmission speed, so that the frequency of the sound wave can shift in the low-frequency direction. In this case, since the frequency characteristic of the output sound wave may change, there may be an effect when detecting intrusion using sound waves. In this way, in order to prevent false detection due to environmental factors such as temperature change or air flow in the security space (s), the motion detection device 100 according to an embodiment of the present invention introduces an adaptive threshold. can do. The adaptive threshold will be described in detail later.

Meanwhile, according to another embodiment, the sound wave generator 110 may preset a waveform in a time domain of the sound wave to be output. That is, the sound wave generator 110 may set and utilize an optimal sound wave capable of easily detecting an event such as an intrusion by reflecting the internal environment or characteristics of the security space (s). In this case, not a specific fixed sound wave, but a sound wave suitable for each security space (s) can be generated and used, so it is possible to optimize the reception performance of sound waves for each of the security spaces (s).

Specifically, the sound wave generator 110 may set a waveform of a sound wave using a time-reverse electric method. Here, the time-reverse electric method corresponds to a technique of receiving a reverberation signal, which is an impulse response to a short-wave signal output in the secure space (s), and then generating a new sound wave obtained by time-reversing the waveform of the reverberant signal and retransmitting it.

Referring to FIG. 3, the shortwave signal may be preset as shown in FIG. 3(a). The short-wave signal may be output in the secure space (s), and as shown in FIG. 3(b), the short-wave signal reflected in the secure space (s) may be received as a reverberation signal.

Thereafter, when the received reverberation signal is matched and filtered, it may appear as shown in Fig. 4(a), and the waveform of the sound wave can be set as shown in Fig. 4(b) by using this. That is, the sound wave generator 110 may set the waveform of the sound wave by rearranging the waveform of the received reverberation signal in the reverse order of time.

The sound waves output in the secure space (s) are reflected by objects included in the secure space (s) and go through different paths, so that they can be received with a time difference. In this case, when a sound wave obtained by time-reversing the waveform of the reverberation signal is output, the signal of the shortest path may be received later, and the signal of the longest path may be received first. That is, the time distribution due to the existing time difference is invalidated, and all of them can reach the receiving point at the same time point. Therefore, the sound waves generated by the time-reverse electric method can be focused on energy by correlation with the security space (s), thereby improving reception performance such as Signal to Noise Ratio (SNR).

In addition, when a single sine wave or multi-tone sound wave is used, the sound field change at a close distance can be detected, but it may be difficult to detect the sound wave diffracted and transmitted in the space behind the object or the change due to a distant or minute movement. have. However, if the time-reverse electric method is used, the environment or characteristics of the security space (s) can be grasped from the reverberation signal, so the energy of the signals that are diffracted by the characteristics of the security space (s) or reflected from a distance It is possible to carry out the focus. Accordingly, the sound wave receiving unit 120 may receive sound waves having a relatively high reception level.

In addition, when the time reversal-based sound waves are used, the difference between the sound wave signals in the case where environmental changes such as temperature or air flow in the security space (s) occur, and the case where there is no change in the environment becomes larger. Therefore, when the sound wave is output using the time-reverse electric method, the detection performance can be improved compared to the case of using the fixed sound wave.

Additionally, the sound wave generator 110 may change the setting timing of the sound wave signal according to the embodiment. That is, according to an embodiment, the sound wave generator 110 may output the first short wave signal in the secure space (s) to set the waveform of the sound wave signal, and then repeatedly output the same sound wave signal.

In addition, according to another embodiment, the sound wave generator 110 may reset and output the waveform of the sound wave signal every period or a specific period. For example, after receiving the reverberation signal of the sound wave output to the secure space (s) in the previous period, the sound wave generator 110 can time-reverse the waveform of the corresponding reverberation signal and reset it to the waveform of the sound wave of the next period. have. In this case, even if an environment change occurs in the security space (s), it is possible to continuously output sound waves optimized for the security space (s). In this case, instead of resetting the waveform of the sound wave every period, the sound wave generator 110 may reset the sound wave whenever a specific number or a specific multiple is reached.

On the other hand, here, the sound wave generator 110 proposes a method of setting the waveform of the sound wave using a time reversal technique, but in addition, if it is possible to automatically generate sound waves suitable for each security space (s), what method Is also applicable.

The sound wave receiver 120 may receive sound waves reflected in the secure space (s), and may generate a plurality of sound wave signals for each frequency using the received sound waves. Here, the sound wave receiving unit 120 may generate a sound wave signal corresponding to the received sound wave to facilitate analysis. Specifically, the sound wave receiver 120 may perform Fast Fourier Transform (FFT) on the received sound wave, and may generate a plurality of sound wave signals for each frequency using the FFT.

Since the sound wave generating unit 110 periodically outputs sound waves, the sound wave receiving unit 120 may generate a plurality of sound wave signals for each frequency corresponding to each sound wave at each time point of receiving the sound wave. Here, the sound wave signal may include each frequency and an amplitude value at the corresponding frequency. According to an embodiment, it is possible to generate sound wave signals for 17 different frequencies every period.

The initial setting unit 130 may set the motion sensing device 100 when the motion sensing device 100 is initially installed. That is, the initial setting unit 130 can automatically analyze the sound wave received by the sound wave receiving unit 120, and the volume of the sound wave output by the sound wave generating unit 110 or the reception of the sound wave receiving unit 120 according to the analysis result You can adjust the sensitivity.

Here, when the initial setting is not performed, a sound wave signal having a negative value or an unstable value may be measured according to the security environment (s) and the setting state in the motion detection device 100. Therefore, by using the initial setting unit 130, the motion detection device 100 can detect the optimal sound wave signal for motion detection in the secure space (s).

For example, the initial setting unit 130 increases the volume of the sound wave or sequentially increases the reception sensitivity of the sound wave receiving unit 120 when the size of the echoed sound wave signal is lower than a predetermined standard to generate a sound wave signal optimized for motion detection. You can get it. In this case, the initial setting unit 130 may automatically adjust and set the sound wave volume or reception sensitivity by software one step at a time. Depending on the embodiment, the reverse case of sequentially lowering the volume and reception sensitivity of sound waves is also possible.

Therefore, by using the initial setting unit 130, the initial setting according to various installation positions or environments of the motion detection device 100 can be easily performed.

The analysis unit 140 may generate analysis data by signal processing the sound wave signals, and may generate an adaptive threshold value reflecting the environmental characteristics of the secure space s.

Here, as one of the signal processing methods, the analysis unit 140 may utilize a cosine similarity between sound wave signals, and at this time, the accumulated value of the cosine similarity between each sound wave signal may be used as analysis data. . That is, when using the cosine similarity, it is possible to analyze a larger amount of data compared to the prior art, and since the accumulated values are used as analysis data, it is possible to increase the accuracy of motion detection.

Specifically, the analysis unit 140 may generate analysis data using the following equation.

Here, CS(t) is the cosine similarity between the respective sound wave signals at time _{t, and f i} (t) and f _j (t) are the i-th and j-th sound wave signals extracted using FFT at time t, respectively. , p is the number of sound wave signals generated at time t, M _i and N _j are the x-coordinate values when representing each sound wave signal as a vector, CS_accum(t) is the cumulative value for the change in cosine similarity, and T is the time point at t. It may be a past point in time to be compared with.

For example, when the sound wave receiver 120 generates 17 sound wave signals every period, two of the 17 sound wave signals may be selected and then cosine similarity for each pair may be calculated. In this case, the cosine similarity for all 136 pairs can be calculated, and then the amount of change in cosine similarity for all 136 pairs (|CS(t)-CS(tT)|) is summed to add the analysis data at time t. Can be generated.

Here, since the analysis unit 140 generates analysis data by using the cosine similarity of each of the sound wave signals, it is possible to more rapidly reflect a change in the sound wave signal that occurs in case of intrusion or the like. On the other hand, in the case of a continuous change of the sound wave signal according to environmental changes such as air inflow in the security space (s), a relatively small amount of change may be reflected. Therefore, by using the analysis data, it is possible to clearly distinguish actions such as intrusion from environmental changes such as air inflow in the security space (s).

On the other hand, in order to determine whether an operation such as an intrusion in the secure space s has occurred at time t using the analysis data, it is necessary to set a reference value. To this end, the analysis unit 140 may generate an adaptive threshold value as a reference value.

Specifically, the analysis unit 140 may set an adaptive threshold value, which is a reference value for determining whether motion is detected in the secure space s, using past data, which is analysis data of sound wave signals received in the past. At this time, the analysis unit 140 is adaptive using environmental adaptation (adaptivity) that is summed by applying weights to each of the past data up to a preset past time point, and sensitivity (Sensitivity) that adjusts the fluctuation range of environmental adaptation. Threshold can be set. That is, instead of simply setting the reference value to one fixed constant, the reference value can be adaptively changed in consideration of changes in the environment over time in the secure space (s). Therefore, it is possible to accurately detect the motion in the secure space (s) in spite of the environmental change in the secure space (s).

First, the analysis unit 140 may calculate an environmental adaptation degree using the following equation. That is, after collecting past data up to a preset past time point, a weight is applied to each of the past data, and the environmental adaptation degree can be calculated by summing them.

Here, S _iN , ..., S _i-1 are determined according to a past time point set as the number of past data, N is the number of past data, and w _n may be a weight applied to each past data. In other words, the analysis data measured before the point t _{can be used as the past data S iN} , ..., S _{i-1 , and the weights w 0} , w ₁ , ..., w _N for each past data can be used. After each setting, the environmental adaptation degree A(S _iN , ..., S _i-1 ) can be calculated by summation. Here, each weight and the number N of past data to be used may be variously set according to embodiments. For example, the more recent past data, the greater the weight, or after obtaining the average value of the past data, a greater weight can be given to the past data close to the average, and each weight allows the operator to derive the optimal result. To do this, it can be set in various ways. Meanwhile, depending on the embodiment, in order to set respective weights, a deep learning technique, a neural network, or the like may be used.

Thereafter, the analysis unit 140 may set an adaptive threshold value using the following equation. That is, a range in which a difference between environmental adaptation and analysis data is allowed can be set as the sensitivity.

Here, A(S _iN , ..., S _i-1 ) is the environmental adaptation level, S _i is the analysis data at the present time, Sensitivity is the sensitivity, and in the case of sensitivity, the operator of the motion detection device 100 is required. It can be arbitrarily set according to. That is, if the current analysis data and environmental adaptation are above the sensitivity, it can be determined that motion has been detected in the secure space (s), and if it is less than the sensitivity, it can be determined that motion in the secure space (s) has not been detected. . Here, the sensitivity corresponds to a factor that adjusts how sensitively the motion detection device 100 performs motion detection.

Additionally, the analysis unit 140 may set an exception situation input by an operator of the motion detection device 100. That is, in addition to determining whether the operation is performed using the adaptive threshold, the operator can set exceptions separately from the adaptive threshold. For example, when a door is opened in the security space (s) or an air conditioner is operated, it is not an intrusion, but a simple environmental change. In this case, it may be necessary to notify for user convenience and the like, and the analysis unit 140 may set reference values for detecting respective exceptions.

In addition, when the analysis data is less than the adaptive threshold, the analysis unit 140 may add the analysis data to the past data. That is, the analysis unit 140 generates each analysis data and an adaptive threshold value in real time. In order to generate an adaptive threshold value in the next period, the analysis data generated in this period may be added to the past data. have. At this time, the oldest data may be deleted and new data may be added. However, as for the added analysis data, only those that are less than the adaptive threshold can be added.

The determination unit 150 may detect an operation in the secure space s by comparing the analysis data with an adaptive threshold value. Specifically, the determination unit 150 may determine one of an event occurrence state in which a preset motion is detected in the secure space (s) and a normal state in which no event occurs in the secure space (s).

That is, if the analysis data is greater than or equal to the adaptive threshold, the determination unit 150 may determine that an event such as an intruder has occurred in the secure space (s). Here, since the determination unit 150 compares the analysis data with the analysis data using an adaptive threshold, it is possible to remove false motion detection according to the environment such as air flow in the secure space (s), and increase the accuracy of motion detection. Do. Thereafter, when it is determined that the event occurs, the motion detection device 100 may notify the occurrence of the event to a preset contact information, or transmit an emergency dispatch request to a police station or a fire station.

On the other hand, if the analysis data is less than the adaptive threshold, the determination unit 150 can determine as a normal state in which no other event has occurred in the secure space (s). You can perform motion detection.

Depending on the embodiment, it is also possible for the determination unit 150 to additionally detect whether the environment has changed in the secure space s, and to determine the environment change state. That is, after determining as a normal state, the determination unit 150 may compare the sound wave signal or analysis data with a preset exception situation, and when a temperature change or abnormal air current in the security space (s) is detected as a result of the comparison, the environment changes. It can be determined by the state. The exception situation may be set in advance by an operator or the like in the analysis unit 140, and the determination unit 150 may use this to determine the environment change state in the secure space (s).

Further, the determination unit 150 may classify each type of environment change corresponding to the environment change state by using the sound wave signal or analysis data. That is, the analysis unit 140 may store sound wave signals, analysis data, etc. corresponding to exceptional situations such as opening a door or operating an air conditioner, and to classify the type of environment change corresponding to each environment change state. can do.

Meanwhile, FIGS. 5 to 8 are graphs showing the performance of the motion detection device according to each embodiment. Specifically, Figs. 5(a) to 8(a) show the performance of the motion detection device (hereinafter, the first motion detection device) according to the first embodiment to which the cosine similarity and the adaptive threshold are applied, and Fig. 5(b) ) To 8(b) show the performance of a motion detection device (hereinafter, a second motion detection device) according to the second embodiment to which a correlation coefficient between sound wave signals and an adaptive threshold value are applied. In addition, Figs. 5(c) to 8(c) show the performance of a motion detection device (hereinafter, a third motion detection device) according to the third embodiment to which a correlation coefficient and a fixed threshold value are applied.

Here, the first motion detection device corresponds to the motion detection device according to an embodiment of the present invention, and the second motion detection device and the third motion detection device correspond to comparison targets thereto.

Meanwhile, the x-axis of each graph corresponds to time, the y-axis corresponds to cosine similarity or correlation coefficient, L1 and L2 denote adaptive threshold values, and L3 denote fixed threshold values. Therefore, referring to the graphs of Figs. 5 to 8, when the cosine similarity or correlation coefficient is greater than or equal to the adaptive threshold (L1, L2) or the fixed threshold (L3), the motion detection device according to each embodiment is It can be determined that an event such as occurrence has occurred. First, FIG. 5 corresponds to a measurement result when an operation such as intrusion occurs in a situation where there is little environmental change in the security space. In this case, the first to third motion detection devices may detect that the cosine similarity or the correlation coefficient exceeds the adaptive threshold values L1 and L2 or the fixed threshold L3 at all similar time points. That is, as shown in Figs. 5(a)(b)(c), it is possible to detect an operation such as an intrusion into a secure space in all three embodiments.

On the other hand, FIGS. 6 and 7 correspond to measurement results when there is no actual intrusion or the like in a situation where environmental changes in the secure space exist. Here, FIG. 7 is a case in which the environmental change is relatively severe compared to FIG. 6.

Referring to FIGS. 6 and 7, the second to third motion sensing devices may detect an erroneous motion due to environmental changes. In particular, it can be seen that, as shown in FIG. 7, when there is a severe environmental change, motion detection using the second to third motion detection devices is virtually impossible. However, as shown in Fig. 7(a), according to the first motion sensing device, it is possible to avoid detecting an erroneous motion in spite of environmental changes.

In addition, FIG. 8 is a measurement result when an operation such as an intrusion (time after 60 sec) occurs in a situation where an environmental change in the security space exists. Here, the intrusion operation corresponds to a part forming a clear peak, as shown on the far right of Figs. 8(a)(b)(c), and the remaining peak noises correspond to environmental changes in the security space. Referring to FIGS. 8(b) and 8(c), it can be seen that the second to third motion detection devices do not distinguish between environmental changes and intrusion actions, and erroneously detect them as intrusion actions. Here, in the case of the second motion detection device, it can be seen that although the motion detection performance is relatively better than that of the third motion detection device, it can be seen that the environment change that appears in 40 to 60 sec is still incorrectly detected as the occurrence of motion. On the other hand, referring to Fig. 8(a), since the first motion detection device can reflect gradually occurring environmental changes using an adaptive threshold, it is possible to accurately detect only intrusion motions despite environmental changes. Do.

9 and 10 are flow charts showing a motion detection method according to an embodiment of the present invention.

9 and 10, the motion detection method according to an embodiment of the present invention includes an initial setting step (S10), a sound wave output step (S20), a sound wave signal reception step (S30), and an analysis data generation step (S40). , It may include an adaptive threshold value generation step (S50) and a motion detection step (S60). Here, each of the steps may be performed by the motion detection device.

Hereinafter, a motion detection method according to an embodiment of the present invention will be described with reference to FIGS. 9 and 10.

In the initial setting step S10, when the motion detection device is initially installed, the motion detection device may be set. That is, the sound wave received by the motion detection device may be automatically analyzed, and the volume of the sound wave to be output or the reception sensitivity of the sound wave to be measured may be adjusted according to the analysis result. At this time, the sound wave volume or reception sensitivity can be set by automatically adjusting one step at a time by software.

Here, depending on the embodiment, it is also possible to preset the waveform of the output sound wave. That is, by reflecting the internal environment or characteristics of the security space, it is possible to set and utilize an optimal sound wave capable of easily detecting an event such as an intrusion. In this case, not a specific fixed sound wave, but a sound wave suitable for each security space can be generated and used, so that the reception performance of sound waves for each security space can be optimized.

Specifically, in the initial setting step (S10), the waveform of the sound wave can be set using the time-reverse electric method, and the sound wave generated by the time-reverse electric method is focused on energy according to the correlation with the secure space, so that the reception performance such as SNR is reduced. Can be improved. That is, after first outputting a short wave signal in the secure space, after confirming the waveform of the received reverberation signal, the waveform of the sound wave can be set by inverting the waveform of the reverberation signal in the time domain.

In the sound wave output step S20, sound waves may be periodically output in the secure space. That is, in order to detect an intrusion or the like in the security space, sound waves may be output in the security space, and sound waves may be periodically output to periodically repeat motion detection in the security space. Depending on the embodiment, sound waves having a frequency in an inaudible frequency band may be used, and the output sound waves may be formed to have a specific pattern by setting the size of each frequency component.

On the other hand, in the sound wave output step (S20), the sound wave set in the initial setting step (S10) may be repeatedly output every cycle. Alternatively, it is possible to reset the sound wave every specific period and output it.

For example, after receiving a reverberation signal of a sound wave output to the secure space in the immediately preceding period, the waveform of the corresponding reverberation signal may be time-reversed and reset to the waveform of the sound wave of the next period. In this case, even if an environment change in the secure space occurs, it is possible to continuously output sound waves optimized for the secure space. Depending on the embodiment, instead of resetting the waveform of the sound wave every period, it is also possible to reset the sound wave every time a specific number or a specific multiple is reached.

In the sound wave signal reception step (S30), a plurality of sound wave signals may be generated for each frequency by measuring sound waves reflected in the secure space. Here, a sound wave signal corresponding to the received sound wave may be generated to facilitate analysis using the received sound wave, and a Fast Fourier Transform (FFT) may be used in some embodiments.

In the analysis data generation step S40, analysis data may be generated by performing signal processing between sound wave signals. Here, as one of the signal processing methods, cosine similarity between sound wave signals can be used, and in this case, a value obtained by accumulating cosine similarity between each sound wave signal can be used as analysis data.

Specifically, in the analysis data generation step S40, analysis data may be generated using the following equation.

Here, CS(t) is the cosine similarity between the respective sound wave signals at time _{t, and f i} (t) and f _j (t) are the i-th and j-th sound wave signals extracted using FFT at time t, respectively. , p is the number of sound wave signals generated at time t, M _i and N _j are the x-coordinate values when representing each sound wave signal as a vector, CS_accum(t) is the cumulative value for the change in cosine similarity, and T is the time point. It may be a point in the past to be compared with. Here, the time point t corresponds to the current time point.

In the adaptive threshold generation step S50, an adaptive threshold value reflecting the environmental characteristics of the secure space may be generated. Specifically, by using past data, which is analysis data of sound wave signals received in the past, an adaptive threshold value, which is a reference value for determining whether motion in a secure space is detected, may be generated. At this time, in the adaptive threshold generation step (S50), the environmental adaptation (adaptivity) that is summed by applying weights to each of the past data up to a preset past time point and the sensitivity (Sensitivity) that adjusts the variation of the environmental adaptation degree are determined. Can be used to set an adaptive threshold. That is, instead of simply setting the reference value to one fixed constant, the reference value can be adaptively changed in consideration of changes in the environment over time in the secure space. Therefore, it is possible to accurately detect motions in the secure space despite changes in the environment within the secure space.

Specifically, in the adaptive threshold generation step (S50), the environmental adaptation degree may be calculated using the following equation. That is, after collecting past data up to a preset past time point, a weight is applied to each of the past data, and the environmental adaptation degree can be calculated by summing them.

Here, S _iN , ..., S _i-1 are determined according to the past time point set as the number of each past data, N is the number of past data, and w _n may be a weight applied to each past data. In other words, analysis data measured before point t _{can be used as past data S iN} , ..., S _i-1 , and weights w0, w1, ..., wN are set for each past data. After that, the environmental adaptation degree A(S _iN , ..., S _i-1 ) can be calculated by summation. Here, each weight and the number N of past data to be used can be set in various ways according to embodiments.

Thereafter, in the adaptive threshold generation step S50, the adaptive threshold may be set using the following equation. That is, a range in which a difference between environmental adaptation and analysis data is allowed can be set as the sensitivity.

Here, A(S _iN , ..., S _i-1 ) is the environmental adaptability, S _i is the analysis data at the present time, and Sensitivity is the sensitivity. Can be set. That is, it is possible to control how sensitively motion detection is performed using the sensitivity.

Additionally, in the adaptive threshold generation step (S50), an exception situation input by the operator of the motion detection device may be set. That is, in addition to the determination of whether to operate using the adaptive threshold, the operator can set cases in which exceptions can be handled separately from the adaptive threshold. In this case, it may be necessary to notify the operator or the like, and reference values for detecting each exception may be set together.

In addition, in the adaptive threshold generation step (S50), when the analysis data is less than the adaptive threshold, the analysis data may be added to the past data. That is, in order to generate an adaptive threshold value in the next period, the analysis data generated in this period may be added to the past data. At this time, the oldest data may be deleted and new data may be added.

In the motion detection step S60, motion detection in the secure space may be performed by comparing the analysis data with an adaptive threshold. Specifically, it can be determined as one of an event occurrence state in which a preset motion in the secure space is detected and a normal state in which no event occurs in the secure space.

That is, referring to FIG. 10, if the analysis data is greater than or equal to the adaptive threshold (S62), it can be determined that an event such as an intruder has occurred in the secure space (S63). Here, since the analysis data is compared with the adaptive threshold value, it is possible to remove false motion detection according to the environment such as air flow in the secure space, and increase the accuracy of motion detection.

On the other hand, if the analysis data is less than the adaptive threshold (S62), it can be determined as a normal state in which no event has occurred in the secure space, and if it is in a normal state, motion detection in the next cycle can be performed without additional measures. have.

On the other hand, depending on the embodiment, it is possible to additionally detect whether the environment has changed in the secure space, and to determine the environment change state. That is, the sound wave signal or analysis data can be compared with a preset exception situation (S64), and when a temperature change or abnormal airflow in the secure space is detected as a result of the comparison, it can be determined as an environment change state (S65). Exception situations can be set in advance by an operator, etc., and can be used to determine the environment change status in the secure space.

Thereafter, when it is determined as the environment change state, each type of environment change corresponding to the environment change state may be classified using sound wave signals or analysis data (S66). In other words, the operator, etc. can store sound wave signals and analysis data corresponding to exceptional situations such as opening a door or operating an air conditioner in advance, so that the types of environment changes corresponding to each environment change state can be classified and provided. have.

On the other hand, if it does not correspond to a preset exception situation (S64), it can be determined as a normal state (S67), and in this case, the analysis data can be reset to the past data (S68). That is, in order to generate an adaptive threshold for the next period, the analysis data may be updated to be added to the past data. Through this, it may not take an additional data reset time for data collection for generating an adaptive threshold.

The present invention described above can be implemented as a computer-readable code on a medium on which a program is recorded. The computer-readable medium may be one that continuously stores a program executable by a computer, or temporarily stores a program for execution or download. In addition, the medium may be a variety of recording means or storage means in a form in which a single piece of hardware or several pieces of hardware are combined, but is not limited to a medium directly connected to a computer system, but may be distributed on a network. Examples of media include magnetic media such as hard disks, floppy disks and magnetic tapes, optical recording media such as CD-ROMs and DVDs, magneto-optical media such as floptical disks, And there may be ones configured to store program instructions, including ROM, RAM, flash memory, and the like. In addition, examples of other media include an app store that distributes applications, a site that supplies or distributes various software, and a recording medium or a storage medium managed by a server. Therefore, the detailed description above should not be construed as restrictive in all respects and should be considered as illustrative. The scope of the present invention should be determined by reasonable interpretation of the appended claims, and all changes within the equivalent scope of the present invention are included in the scope of the present invention.

The present invention is not limited by the above-described embodiments and the accompanying drawings. It will be apparent to those of ordinary skill in the art to which the present invention pertains, that components according to the present invention can be substituted, modified, and changed within the scope of the technical spirit of the present invention.

100: motion detection device 110: sound wave generator
120: sound wave receiving unit 130: initial setting unit
140: analysis unit 150: discrimination unit

Claims (15)

A sound wave generator for periodically outputting sound waves in the secure space;
A sound wave receiver configured to measure sound waves reflected in the secure space and generate a plurality of sound wave signals for each frequency;
An analysis unit for generating analysis data by signal processing the sound wave signals, and generating an adaptive threshold reflecting the environmental characteristics of the secure space; And
A motion detection device comprising a discriminator configured to detect a motion in a secure space by comparing the analysis data with an adaptive threshold.
The method of claim 1, wherein the sound wave generator
And setting a waveform of the sound wave using the measured reverberation signal after outputting a short wave signal into the secure space.
The method of claim 2, wherein the sound wave generator
And resetting a waveform of a sound wave of a next period by using the reverberation signal of the sound wave output in the secure space.
The method of claim 2, wherein the sound wave generator
And rearranging the waveforms of the reverberant signals in the reverse order of time and setting them as the waveforms of the sound waves.
The method of claim 1, wherein the sound wave receiver
A motion detection device, characterized in that by performing a Fast Fourier Transform (FFT), generating a plurality of sound wave signals for each frequency of each sound wave.
The method of claim 1,
And an initial setting unit configured to analyze the sound wave received by the sound wave receiving unit and adjust the volume of the sound wave output by the sound wave generating unit or a reception sensitivity of the sound wave receiving unit.
The method of claim 1, wherein the analysis unit
And generating the analysis data by calculating a cosine similarity between the sound wave signals.
The method of claim 1, wherein the adaptive threshold is
A motion sensing device, characterized in that it is set using an environmental adaptation level based on past data, which is analysis data of sound wave signals received in the past, and a sensitivity that adjusts a fluctuation range of the environmental adaptation level.
The method of claim 8, wherein the analysis unit
A motion detection apparatus comprising: collecting past data up to a preset past time point, applying a weight to each of the past data, and summing them to calculate the environmental adaptability.
The method of claim 9, wherein the analysis unit
And a range in which a difference between the environmental adaptation and the analysis data is allowed is set as the sensitivity.
The method of claim 1, wherein the determination unit
And determining one of an event occurrence state in which a preset motion is detected in the secure space and a normal state in which no event occurs in the secure space.
The method of claim 11, wherein the determination unit
When it is determined as the normal state, by comparing the sound wave signal or analysis data with a preset exception situation, the motion detection device, characterized in that it determines whether it corresponds to an environment change state corresponding to the occurrence of temperature change or abnormal air current in the security space. .
The method of claim 12, wherein the determination unit
And classifying each type of environment change corresponding to the environment change state by using the sound wave signal or analysis data.
The method of claim 11, wherein the analysis unit
When it is determined as the normal state, analysis data corresponding to the normal state is added to the past data,
The adaptive threshold is
A motion sensing device, characterized in that it is set using an environmental adaptation based on past data, which is analysis data of sound wave signals received in the past, and a sensitivity for adjusting a variation range of the environmental adaptation.
In the motion detection method in a secure space using a motion detection device,
Periodically outputting sound waves in the secure space;
Measuring sound waves reflected in the secure space and generating a plurality of sound wave signals for each frequency;
Generating analysis data by performing signal processing between the sound wave signals;
Generating an adaptive threshold reflecting environmental characteristics of the secure space; And
And performing motion detection in a secure space by comparing the analysis data with an adaptive threshold.

Images