TW201832183A - Vibration detection device, vibration detection method and vibration detection program - Google Patents

Abstract

In the present invention, a normal feature storage unit (131) stores normal feature information (301) representing a normal vibration feature. A histogram information generation unit (110) acquires moving image data (109), and extracts from the moving image data a motion vector representing the motion of a vibrating object. In addition, the histogram information generation unit (110) counts the quantity of motion vectors for each motion direction range, and uses the motion vector quantity for each motion direction range to generate histogram information (21) for each frame of the moving image data (109). In addition, a frequency analysis unit (122) uses the histogram information (21) to calculate, as vibration feature information (302), the frequencies of time-series changes in the motion vector quantity for each motion direction range, and the frequency components for each frequency. A comparator compares the vibration feature information (302) and the normal feature information (301).

Description

 Vibration detecting device, vibration detecting method, and vibration detecting program product  

The present invention relates to a vibration detecting device, a vibration detecting method, and a vibration detecting program product.

In the image vibration detection method disclosed in Patent Document 1, in the captured image captured by the imaging unit, a moving component that moves up and down with vibration in the image is specified, and The movement history of the specified moving component is subjected to frequency analysis, and the intensity of the vibration frequency of the moving component is calculated. Next, the vibration detecting method extracts a vibration frequency of a predetermined range among the calculated vibration frequency of the moving component. Next, the vibration detecting method extracts the captured image captured by the balance point of the vibration within the predetermined frequency range as a balanced image based on the frequency intensity in the extracted predetermined frequency range. Next, the vibration detecting method calculates the image speed in the balanced image, and uses the image speed in the calculated balanced image to exclude the image speed due to the stable vibration from the captured image. Next, the vibration detecting method detects the center position of the displacement in the image displaced by the vibration based on the captured image after the image speed due to the stable vibration is excluded. Further, the vibration detecting method monitors the center position of the detected displacement and calculates the amount of displacement of the image due to the vibration of the image captured by the imaging unit.

[Previous Technical Literature]

[Patent Literature]

Patent Document 1: Japanese Laid-Open Patent Publication No. 2006-170961

According to the technique disclosed in Patent Document 1, it is known in advance that the moving component having a concentration gradient in the vertical direction is vibrated in the vertical direction of the screen, and there is a problem that the vibration cannot be applied to any direction of the screen. Further, the technique disclosed in Patent Document 1 has a problem of detecting vibration only, and there is a problem that it is impossible to determine whether the vibration is normal vibration or abnormal vibration and abnormal vibration is detected.

An object of the present invention is to detect a vibration state from a video regardless of a direction in which a vibration detecting object vibrates, and to determine whether it is a normal vibration or an abnormal vibration from a difference in a vibration state, and to detect an abnormal vibration of a vibration detecting target.

The vibration detecting device of the present invention includes: a normal feature memory portion that memorizes normal feature information indicating a characteristic of normal vibration among vibrations of a vibrating object to be vibrated; a histogram information generating portion that obtains a representation A moving image data of the moving image of the vibrating object is captured, and an action vector indicating an action of the vibrating object is extracted from the moving image data, and each action direction range depending on a range of directions of the motion vector Counting the number of motion vectors, and using the number of motion vectors in each of the foregoing motion direction ranges, generating a histogram of each frame of the moving image data as histogram information; a frequency analysis unit Using the aforementioned histogram information, calculating the frequency of the time series change of the number of motion vectors in each of the aforementioned motion direction ranges and the frequency component of each frequency as the vibration characteristic information indicating the characteristics of the vibration of the vibrating object; and comparing a portion that obtains the normal feature information from the normal feature memory portion and compares the vibration Normal sign information and said feature information, and determines whether the feature information having the vibration characteristics of normal information.

In the vibration detecting device of the present invention, the normal feature memory unit stores normal feature information indicating the characteristics of the normal vibration among the vibrations of the vibrating object to be vibrated. The histogram information generating unit acquires moving image data indicating a moving image of the vibrating object, and extracts an motion vector indicating the motion of the vibrating object from the moving image data. Further, the histogram information generating unit counts the number of motion vectors according to each motion direction range of the range of the direction of the motion vector, and generates each of the motion image data using the number of motion vectors of each motion direction range. The histogram of the packet is used as histogram information. Further, the frequency analysis unit calculates the frequency of the time series change of the number of motion vectors in each direction of the motion direction and the frequency component of each frequency using the histogram information as the vibration characteristic information indicating the characteristics of the vibration of the vibrating object. . Furthermore, the comparison department compares the vibration feature information with the normal feature information, and determines whether the vibration feature information has normal feature information. Therefore, according to the vibration detecting device of the present invention, it is possible to detect the vibration state from the image regardless of the direction in which the vibration detecting object vibrates, and thereby determine whether the vibration is abnormal or abnormal vibration from the difference in the vibration state, and detect the abnormal vibration of the vibration detecting object. .

11‧‧‧Dynamic image compression data

21‧‧‧Histogram Information

31‧‧‧Dynamic image data package

109‧‧‧Dynamic image data

100, 100a‧‧‧ vibration detecting device

110, 110a‧‧‧ Histogram Information Generation Department

111‧‧‧Compressed Data Receiving Department

112‧‧‧Action Vector Extraction Department

113‧‧‧Action direction calculation department

114‧‧‧Action Counting Department

115‧‧‧Action histogram generation department

120‧‧‧Vibration Analysis Department

121‧‧‧Action Quantity Extraction Department

122‧‧‧Frequency Analysis Department

123‧‧‧Normal feature extraction department

124‧‧‧Comparative Department

125‧‧‧Notice Department

130‧‧‧Memory Department

131‧‧‧Normal Characteristic Memory

301‧‧‧Normal characteristics information

302‧‧‧Vibration characteristics information

303‧‧‧Comparative results

311‧‧‧Data packet receiving department

312‧‧‧moving body

313‧‧‧Dynamic motion vector extraction

501‧‧‧Dynamic image data package

502‧‧‧ vibrating objects

503‧‧‧Vibration

504, 505‧‧‧ cameras

506‧‧‧Action Vector

610‧‧‧Vibration detection method

620‧‧‧Vibration detection program

909‧‧‧Processing Circuit

910‧‧‧ processor

920‧‧‧ memory device

921‧‧‧ memory

922‧‧‧Auxiliary memory device

930‧‧‧Input interface

940‧‧‧output interface

S100‧‧‧Vibration detection processing

S10‧‧‧Normal vibration extraction treatment

S20‧‧‧Vibration feature comparison processing

S110, S110a‧‧‧ Histogram information generation processing

S120‧‧‧ frequency analysis processing

S130‧‧‧Normal feature extraction processing

S140‧‧‧Comparative processing

S220‧‧‧Exception determination processing

Fig. 1 is a configuration diagram of a vibration detecting device 100 according to the first embodiment.

Fig. 2 is a view for explaining the moving image compression material 11 of the first embodiment.

Fig. 3 is a flowchart showing the vibration detecting method 610 of the vibration detecting method 610 and the vibration detecting program 620 of the first embodiment.

Fig. 4 is a flowchart showing the normal vibration extraction processing S10 of the first embodiment.

Fig. 5 is a frequency component of each frequency obtained by performing frequency analysis on the number of motion vectors in a specific operation direction range from the motion vector extracted from the image of the vibrating object 502 captured in the first embodiment. An example of chronological ordering.

Fig. 6 is a flowchart showing the vibration characteristic comparison processing S20 of the first embodiment.

Fig. 7 is a frequency component of each frequency obtained by performing frequency analysis on the number of motion vectors in a specific operation direction range from the motion vector extracted from the image of the vibrating object 502 captured in the first embodiment. An example of chronological ordering.

Fig. 8 is a configuration diagram of a vibration detecting device 100 according to a modification of the first embodiment.

Fig. 9 is a configuration diagram of the vibration detecting device 100a of the second embodiment.

Fig. 10 is a flowchart showing the histogram information generation processing S110a of the second embodiment.

Hereinafter, embodiments of the present invention will be described using the drawings. In the drawings, the same reference numerals are given to the same or equivalent parts. In the description of the embodiments, the same or equivalent portions will be appropriately omitted or simplified.

Embodiment 1

***Composed description***

The configuration of the vibration detecting device 100 of the present embodiment will be described using Fig. 1 .

As shown in Fig. 1, the vibration detecting device 100 is a computer.

The vibration detecting device 100 includes a processor 910, a memory device 920, an input interface 930, and an output interface 940. The memory device 920 includes a memory 921 and an auxiliary memory device 922.

The vibration detecting device 100 includes a histogram information generating unit 110, a vibration analyzing unit 120, and a memory unit 130 as functional configurations.

The histogram information generating unit 110 includes a compressed data receiving unit 111, an operation vector extracting unit 112, an operation direction calculating unit 113, an operation number counting unit 114, and an operation histogram generating unit 115.

The vibration analysis unit 120 includes an operation number extraction unit 121, a frequency analysis unit 122, a normal feature extraction unit 123, a comparison unit 124, and a notification unit 125.

The memory unit 130 includes a normal feature storage unit 131. The normal feature storage unit 131 is for storing the normal feature information 301 indicating the characteristics of the normal vibration among the vibrations of the vibrating object to be vibrated. Further, in the memory unit 130, the histogram information 21 is stored. The motion number extracting unit 121 obtains the histogram information 21, and extracts the number of motion vectors for each motion direction range.

The compressed data receiving unit 111, the motion vector extracting unit 112, the motion direction calculating unit 113, the motion number counting unit 114, the motion histogram generating unit 115, the motion number extracting unit 121, the frequency analyzing unit 122, the normal feature extracting unit 123, and the comparing unit The functions of each of the 124 and the notification unit 125 are implemented by software. In the following description, the compressed data receiving unit 111, the motion vector extracting unit 112, the operation direction calculating unit 113, the motion number counting unit 114, the motion histogram generating unit 115, the motion number extracting unit 121, and the frequency analyzing unit 122 are provided. The normal feature extracting unit 123, the comparing unit 124, and the notifying unit 125 are referred to as respective parts of the histogram information generating unit 110 and the vibration analyzing unit 120.

The memory unit 130 is realized by the memory 921. Further, the memory unit 130 can be realized only by the auxiliary memory device 922 or by the memory 921 and the auxiliary memory device 922. The method of implementing the memory unit 130 is arbitrary.

The processor 910 is connected to other hardware through a signal line for controlling these other hardware. The processor 910 is an IC (Integrated Circuit) that performs arithmetic processing. A specific example of the processor 910 is a CPU (Central Processing Unit), a DSP (Digital Signal Processor), and a GPU (Graphics Processing Unit).

Memory 921 is a memory device for temporary memory data. Specific examples of the memory 921 are SRAM (Static Random Access Memory) and DRAM (Dynamic Random Access Memory).

The auxiliary memory device 922 is a memory device that stores data. A specific example of the auxiliary memory device 922 is an HDD (Hard Disk Drive). In addition, the auxiliary memory device 922 can be an SD (registered trademark) (Secure Digital) memory card, a CF (CompactFlash), a NAND (reverse) flash, a flexible disk (flexible disk). ), portable optical media such as optical disk, CD (compact disk), blue-ray (registered trademark) disk, DVD (Digital Versatile Disc).

The input interface 930 is a port to which an imaging device such as a camera 504 is connected. The input interface 930 acquires the moving image compressed material 11 indicating the moving image captured by the camera 504, and outputs it to the compressed data receiving unit 111. Alternatively, the input interface 930 is a port that is connected to an input device such as a mouse, a keyboard, or a touch panel. Specifically, the input interface 930 is a USB (Universal Serial Bus) terminal. In addition, the input interface 930 may be a port connected to a LAN (Local Area Network).

The output interface 940 is a port for a cable connection of a display device such as a display. Specifically, the output interface 940 is a USB terminal or an HDMI (registered trademark) (High Definition Multimedia Interface) terminal. Specifically, the display is an LCD (Liquid Crystal Display).

In the auxiliary memory device 922, a program for realizing the functions of each of the histogram information generating unit 110 and the vibration analyzing unit 120 is stored. The program for realizing the functions of each of the histogram information generating unit 110 and the vibration analyzing unit 120 is also referred to as a vibration detecting program 620. This program is loaded into the memory 921 and read to the processor 910 for execution by the processor 910. Further, the auxiliary memory device 922 stores an OS (Operating System). At least a portion of the OS of the auxiliary memory device 922 is loaded in the memory 921. The processor 910 executes the vibration detecting program 620 while executing the OS.

The vibration detecting device 100 may include only one processor 910, and may also include a plurality of processors 910. The plurality of processors 910 can also be executed in cooperation with a program that implements functions of each of the histogram information generating unit 110 and the vibration analyzing unit 120.

Information, data, signal values, and variable values indicating the results of processing by each of the histogram information generating unit 110 and the vibration analyzing unit 120 are stored in the auxiliary storage device 922, the memory 921, or the processing of the vibration detecting device 100. A register or flash memory in the device 910.

A program for realizing the functions of each of the histogram information generating unit 110 and the vibration analyzing unit 120 can be stored in the portable recording medium. The portable recording medium is specifically a magnetic disk, a flexible disk, a magnetic disk, a compact disk (CD), a blue-ray (registered trademark) disk, and a DVD (Digital). A memory card such as a Versatile Disc, a digital versatile disc, or an SD (registered trademark) card.

Further, the vibration detecting program product is a memory medium and a memory device in which the vibration detecting program 620 is recorded. The vibration test program product refers to a program that is loaded with a computer readable program regardless of its appearance.

The moving image compression material 11 of the present embodiment will be described using Fig. 2 . In Fig. 2, a vibrating object 502 in the moving image packet 501, a camera 504 that images the vibrating object, and an example of the moving image compressed material 11 output from the camera 504 are displayed. The moving image compression data 11 is an example of moving image data 109 indicating a moving image of the vibrating object 502.

The moving image data packet 501 shown in FIG. 2 is a packet in which the image of the vibrating object 502 is captured by the camera 504. The vibration 503 of the vibrating object 502 is represented by a slight change in the position of the vibrating object 502 between the moving image data packets 501.

The camera 504 of the present embodiment includes a moving image compression function for recording a small change in the position of the vibrating object 502 due to the vibration 503 as the motion vector 506 in the moving image compressed material 11. The motion vector 506 represents a slight change in the position of the vibrating object 502 between the moving image data packets 501 by a vector.

*** Description of action***

The vibration detecting process 610 of the vibration detecting method 610 and the vibration detecting program 620 of the present embodiment will be described with reference to Fig. 3 .

The vibration detecting process S100 includes a normal vibration extracting process S10 and a vibration feature comparing process S20.

<Normal vibration extraction processing S10>

The normal vibration extraction process S10 of the present embodiment will be described using Fig. 4 .

The normal vibration extraction processing S10 is a process of extracting the vibration characteristics of the vibration 502 of the normal vibration from the moving image compression data 11 as the normal feature information 301 and memorizing it in the normal feature storage unit 131. The normal vibration extraction processing S10 includes a histogram information generation processing S110, a frequency analysis processing S120, and a normal feature extraction processing S130.

<Histogram information generation processing S110>

In the histogram information generation processing S110, the histogram information generation unit 110 acquires the motion image data 109 indicating the moving image of the vibrating object 502, and extracts the motion indicating the motion of the vibrating object 502 from the moving image data 109. Vector 506. Furthermore, the histogram information generating unit 110 counts the number of motion vectors for each range of the operation direction in the range of the direction of the motion vector. Furthermore, the histogram information generating unit 110 generates a histogram of each packet of the moving image data 109 as the histogram information 21 using the number of motion vectors in each motion direction range. Further, the histogram information generating unit 110 acquires the moving image compressed material 11 including the compressed motion image and is the moving image data 11 including the motion vector 506 as the moving image data 109. Further, the histogram information generating unit 110 extracts the motion vector 506 from the moving image compressed material 11.

Specifically, first, in step S111, the vibrating object 502 is imaged by the camera 504 having the dynamic image compression function.

In step S112, the compressed data receiving unit 111 outputs the moving image compressed material 11 output from the camera 504. The compressed data receiving unit 111 outputs the moving image compressed data of one moving image data packet. Here, in the moving image compression data 11, an action vector 506 calculated from the block information of the information between the close dynamic image data packets 501 is included. By way of a specific example, the motion vector 506 is an encoding motion vector defined by a compression method such as an MPEG (Moving Picture Expert Group) format. This coded motion vector is calculated in units of pixel blocks from information such as the brightness gradient between the moving image packets 501.

In step S113, the motion vector extracting unit 112 extracts all the motion vectors 506 included in each of the moving image data packets 501 from the moving image compressed data 11 received by the compressed data receiving unit 111.

In step S114, the operation direction calculation unit 113 acquires the motion vector 506 from the motion vector extraction unit 112. The operation direction calculation unit 113 calculates the arctangent of the X component and the Y component for each motion vector, thereby obtaining the operation direction of the motion vector. Further, the operation direction calculation unit 113 calculates the magnitude of the motion vector by calculating the square root of the sum of the square of the X component and the square of the Y component for each motion vector.

In step S115, the operation number counting unit 114 acquires the operation direction and size of each motion vector obtained by the operation direction calculating unit 113. The motion number counting unit 114 determines which one of the motion direction ranges that fall within the range of the motion vector is the motion direction of the motion vector, and classifies the motion vector by each motion direction range. The range of the operation direction refers to each angular range in which 360° is divided into m equal parts. The number-of-operations counting unit 114 classifies the motion vector into the motion direction range 1, the motion direction range 2, and the motion direction range 2, depending on which range of motion directions each of the motion vectors extracted by the motion vector extraction unit 112 falls into, and 动作 ̇ ̇ The range of motion directions is m. The motion number counting unit 114 counts the number of motion vectors classified in the motion direction range for each motion direction range. The number-of-actions counting unit 114 can count only the number of motion vectors regardless of the size of the motion vector, and can also count the number according to the magnitude of the motion vector. Further, the number-of-actions counting unit 114 may count only when the magnitude of the motion vector is equal to or greater than a certain value. At this time, a threshold for determining whether or not the magnitude of the motion vector is equal to or greater than a certain value may be provided.

In step S116, the operation histogram generation unit 115 acquires the number of motion vectors counted for each operation direction range from the operation number counting unit 114. The motion histogram generation unit 115 generates a histogram of each packet of the moving image data 109 using the number of motion vectors in each motion direction range. Specifically, the motion histogram generation unit 115 generates a histogram of one moving image data packet in which the horizontal axis is the number 1 to m of the operation direction range, and the vertical axis is the number of motion vectors. The histogram generating unit 115 stores the histogram of the generated 1 moving image data packet in the memory unit 130 as the histogram information 21. In the histogram information 21 of the memory unit 130, a histogram generated by each of the moving image data packets constituting the moving image is stored in time series.

<Frequency analysis processing S120>

In the frequency analysis processing S120, the frequency analysis unit 122 calculates the frequency of the time series change of the number of motion vectors in each of the operation direction ranges and the frequency component of each frequency using the histogram information 21 as the vibration object 502. Vibration characteristic information 302 of the characteristics of the vibration. Specifically, the frequency analysis unit 122 performs a Fourier transform on the time series change of the number of motion vectors in each motion direction range, thereby calculating a time series change of the number of motion vectors in each motion direction range. The frequency and the frequency component of each frequency are used as the vibration characteristic information 302.

More specifically, first, in step S121, the motion number extracting unit 121 extracts the latest histogram information 21 of the number of consecutive moving image data packets (=n) required for frequency analysis from the histogram information 21. And extracting the number of motion vectors of n packets of packets in a range of each action direction (1 to m).

In step S122, the frequency analysis unit 122 acquires the number of motion vectors of consecutive n packets of each operation direction range from the operation number extracting unit 121. The frequency analysis unit 122 performs Fourier transform on the time series change of the number of motion vectors in each operation direction range, and obtains the frequency component of each frequency. Furthermore, the frequency analysis unit 122 outputs the frequency of the time series change of the number of motion vectors and the frequency component of each frequency as the vibration characteristic information 302 in each of the operation direction ranges.

<Normal feature extraction processing S130>

In the normal feature extraction processing S130, the normal feature extraction unit 123 extracts information indicating the characteristics of the normal vibration from the vibration characteristic information 302 calculated by the frequency analysis unit 122 as the normal feature information 301.

Specifically, first, in step S123, the normal feature extracting unit 123 acquires the frequency of the time series change of the number of motion vectors in each of the operating direction ranges and the frequency component of each frequency from the frequency analyzing unit 122. In other words, the normal feature extracting unit 123 acquires the vibration characteristic information 302 calculated by the frequency analyzing unit 122. The normal feature extracting unit 123 stores the frequency components of each frequency of the time series change of the number of motion vectors in each of the operation direction ranges in time series. The normal feature extracting unit 123 extracts the characteristics of the vibration of the vibrating object 502 of the normal vibration from the frequency component of each frequency of the time series of the number of motion vectors in each of the operating direction ranges stored in time series as the normal feature information. 301.

In step S123, the normal feature extracting unit 123 stores the normal feature information 301 in the normal feature storage unit 131 of the storage unit 130.

Fig. 5 is a view showing frequency components of each frequency obtained by performing frequency analysis on the number of motion vectors in a specific operation direction range in the motion vector extracted from the image of the vibrating object 502 captured in the first embodiment. An example of chronological ordering.

Fig. 5 is a view showing a state in which a frequency component of a specific frequency of a time series change of the number of motion vectors is larger than a frequency component of other frequencies in a specific range of motion directions, and is maintained even if time changes. An example of the characteristics of normal vibration. The normal vibration characteristic may also be a plurality of combinations of the magnitudes of the frequency components of each frequency of the time series change of the number of motion vectors in each of the motion direction ranges of the motion vector extracted from the motion image compression data. Further, a pattern of the time series change may be presented, specifically, a mode in which a frequency component of a specific frequency is periodically increased or decreased, or a mode in which a frequency component is periodically changed from a frequency having a larger frequency than other frequencies.

<Vibration Feature Comparison Processing S20>

The vibration characteristic comparison processing S20 of the present embodiment will be described using Fig. 6 .

The vibration characteristic comparison processing S20 is processing for determining whether the vibration of the vibrating object 502 obtained from the moving image compression data 11 is normal vibration or abnormal vibration, and notifying the determination result.

The vibration characteristic comparison processing S20 includes a histogram information generation processing S110 and an abnormality determination processing S220. The histogram information generation processing S110 is the same processing as the histogram information generation processing S110 described in the normal vibration extraction processing S10.

<Histogram information generation processing S110>

First, the histogram information generation processing S110 executes the histogram information generation processing S110.

Steps S111 to S116 of Fig. 6 are the same as steps S111 to S116 of Fig. 4 .

<Frequency analysis processing S120>

Steps S121 to S122 of Fig. 6 are the same as steps S121 to S122 of Fig. 4 .

<Comparative Processing S140>

The comparison unit 124 obtains the normal feature information 301 from the normal feature storage unit 131, and compares the vibration feature information 302 with the normal feature information 301 to determine whether the vibration feature information 302 has the normal feature information 301. Further, when the comparison unit 124 determines that the vibration characteristic information 302 does not have the normal feature information 301, the notification unit 125 outputs an abnormal vibration notification for notifying the abnormality of the vibration. Further, when the comparison unit 124 determines that the vibration characteristic information 302 has the normal feature information 301, the notification unit 125 outputs a normal vibration notification informing that the vibration is normal.

Specifically, first, in step S223, the comparison unit 124 acquires the frequency component of each frequency of the time series change of the number of motion vectors in each of the operation direction ranges from the frequency analysis unit 122. In other words, the comparison unit 124 acquires the vibration characteristic information 302 calculated by the frequency analysis unit 122. The comparison unit 124 compares the vibration characteristic information 302 with the normal feature information 301 stored in the normal feature storage unit 131. The comparing unit 124 compares whether or not there is a feature corresponding to the normal feature information 301 among the frequency components of each of the frequency-series changes in the number of motion vectors in each of the motion direction ranges stored in the time series. The comparison unit 124 may correct the range of the operation direction in consideration of the possibility that the tilt of the vibrating object during photographing deviates from the tilt of the vibrating object during normal vibration extraction.

The comparison unit 124 outputs a comparison result of the vibration of the vibrating object 502 in the comparative imaging and the characteristic of the normal vibration.

Fig. 7 is a frequency component of each frequency obtained by performing frequency analysis on the number of motion vectors in a specific operation direction range from the motion vector extracted from the image of the vibrating object 502 captured in the present embodiment. A diagram of an example of a time series arrangement.

Fig. 7 is an example of a time-series change in the frequency and frequency components of the time-series change in the number of motion vectors in the range of the operation direction in the same range as the operation direction range illustrated in Fig. 5. Among the features of the normal vibration exemplified in Fig. 5, there are more frequency components in the specific frequency than the other frequencies, even if the time changes. However, in the time series variation of the frequency and frequency components exemplified in Fig. 7, there is no such characteristic of normal vibration. Therefore, the comparison unit 124 determines that the vibration is a characteristic having a feature different from the normal vibration.

The comparison unit 124 sets the normal vibration detection when the frequency component of each frequency of the time series change of the number of motion vectors in each of the motion direction ranges stored in the time series has a feature consistent with the normal feature information 301. As a result of the comparison 303. Further, the comparison unit 124 sets an abnormality when the frequency component of each frequency of the time series change of the number of motion vectors in each of the motion direction ranges stored in the time series does not have the feature consistent with the normal feature information 301. The vibration is detected as a comparison result 303.

In step S224, the notification unit 125 determines whether the comparison result 303 is normal vibration detection or abnormal vibration detection. In the case of normal vibration detection, the process proceeds to step S243. In the case of abnormal vibration detection, the process proceeds to step S244.

In step S225, the notification unit 125 notifies the normal vibration detection.

In step S226, the notification unit 125 notifies the abnormal vibration detection.

***Other composition***

Further, the vibration detecting device 100 of the present embodiment may have a communication device that communicates with other devices through a network. The communication device has a receiver and a transmitter. The communication device is connected to a communication network such as a LAN, an internet, or a telephone line by wire or wirelessly. Specifically, the communication device is a communication chip or a NIC (Network Interface Card). The communication device is a communication unit that communicates data. The receiver is a receiving unit that receives data. The transmitter is a transmission unit that transmits data. The vibration detecting device 100 of the present embodiment can also obtain an operation description and a non-functional requirement through the communication device. Further, the vibration detecting device 100 of the present embodiment can transmit the switching operation description to another device via the communication device.

In the present embodiment, the functions of the respective parts of the histogram information generating unit 110 and the vibration analyzing unit 120 are realized by software. However, as a modification, the functions of the respective portions of the histogram information generating unit 110 and the vibration analyzing unit 120 may be realized by hardware.

The configuration of the vibration detecting device 100 according to the modification of the embodiment will be described with reference to Fig. 8.

As shown in FIG. 8, the vibration detecting device 100 includes a processing circuit 909, an input interface 930, and an output interface 940.

The processing circuit 909 is a dedicated electronic circuit that realizes the functions of the respective portions of the histogram information generating unit 110 and the vibration analyzing unit 120 described above and the memory unit 130. Specifically, the processing circuit 909 is a single circuit, a composite circuit, a programmed processor, a parallel programmed processor, a logic IC, a GA, an ASIC, or an FPGA. GA is an abbreviation for Gate Array. The ASIC is an abbreviation for Application Specific Integrated Circuits. The FPGA is an abbreviation for Field-Programmable Gate Array.

The functions of the respective portions of the histogram information generating unit 110 and the vibration analyzing unit 120 can be realized by one processing circuit 909, or can be realized by being distributed over a plurality of processing circuits 909.

As another modification, the functions of the respective portions of the histogram information generating unit 110 and the vibration analyzing unit 120 can be realized by a combination of a software and a hardware. That is, the function of a part of the vibration detecting device 100 can also be realized by a dedicated hardware, and the remaining functions are realized by software.

The processor 910, the memory device 920, and the processing circuit 909 of the vibration detecting device 100 are collectively referred to as "processing circuitry". In other words, even if the configuration of the vibration detecting device 100 is the configuration shown in any one of the first and eighth figures, the functions of the respective portions of the histogram information generating unit 110 and the vibration analyzing unit 120 and the memory unit 130 are also processed by the processing circuit. The system is implemented.

"Department" may also be referred to as "step" or "procedure" or "handling". In addition, the function of the "part" can also be realized by firmware.

*** Explanation of the effect of this embodiment***

As described above, the vibration detecting device 100 of the present embodiment extracts an operation vector of the vibration from the image of the vibrating object, and performs frequency analysis in each of the operating direction ranges. Further, the vibration detecting apparatus 100 according to the present embodiment extracts normal feature information belonging to the characteristics of the normal vibration in advance based on the time-series change of the result of the frequency analysis. Further, the vibration detecting device 100 of the present embodiment compares the vibration characteristic information of the vibrating object captured by the vibration detecting device 100 with the normal characteristic information.

Therefore, according to the vibration detecting device of the present embodiment, the vibration state can be detected from the image regardless of the direction in which the vibration detecting object vibrates. Further, according to the vibration detecting device of the present embodiment, whether the normal vibration or the abnormal vibration is determined from the difference in the vibration state can be detected to detect the abnormal vibration of the vibration detecting target. Further, according to the vibration detecting device of the present embodiment, whether the normal vibration or the abnormal vibration is determined from the difference in the vibration state can be detected to detect the abnormal vibration of the vibration detecting target. Further, according to the vibration detecting device of the present embodiment, it is possible to determine whether it is normal vibration or abnormal vibration without being affected by the vibration and positional deviation of the camera, the position of the vibrating object, and the deviation of the inclination.

Embodiment 2

In the present embodiment, the difference from the first embodiment will be mainly described.

In the first embodiment, the mode of extracting the motion vector of the moving image compression data 11 has been described. In the present embodiment, when the motion vector is extracted by detecting the movement of the vibrating object 502 between the adjacent moving image packets 31, it is determined whether the vibration of the vibrating object 502 is normal or abnormal.

The configuration of the vibration detecting device 100a of the present embodiment will be described with reference to Fig. 9. In the ninth embodiment, the same components as those described in the first embodiment are denoted by the same reference numerals, and their description will be omitted.

In Fig. 9, the vibration detecting device 100a includes a histogram information generating unit 110a instead of the histogram information generating unit 110 of Fig. 1 . The histogram information generating unit 110a acquires the moving image data packet 31 constituting the moving image as the moving image data 109, and detects the movement of the vibrating object between the adjacent moving image data packets 31, thereby extracting the motion vector. 506. Further, the histogram information generating unit 110a may extract the motion vector 506 by detecting the movement of the vibrating object between the adjacent moving image data packets 31. The adjacent moving image data packet 31 is a concept included in the close moving image data packet 31. The histogram information generating unit 110a includes a packet receiving unit 311 that receives the moving image packet 31, a moving body extracting unit 312, and a dynamic motion vector extracting unit 313 instead of the compressed data receiving unit 111 of Fig. 1 and The motion vector extracting unit 112.

The packet receiving unit 311 receives the moving image packet 31 output from the camera 505 having no moving image compression function through the input interface 930.

The other configuration of the vibration detecting device 100a of the present embodiment is the same as that of the first embodiment.

The histogram information generation processing S110a of this embodiment will be described using Fig. 10 . In the tenth diagram, the points different from the fourth and sixth embodiments described in the first embodiment are the processing of steps S111 to S113.

In step S111a, the vibrator 502 is photographed by the camera 505 which does not have the dynamic image compression function.

In step S112a, the packet receiving unit 311 outputs the moving image packet 31 output from the camera 505.

In step S113a, the dynamic body extracting unit 312 sequentially inputs the moving image data packets from the packet receiving unit 311, and extracts the dynamic body information in the screen from the difference between the adjacent moving image data packets. The moving body motion vector extracting unit 313 inputs the moving body information from the moving body extracting unit 312, and generates an motion vector from the direction of the motion and the magnitude of the motion.

The following operations are the same as those in the first embodiment.

As described above, according to the vibration detecting apparatus 100 of the present embodiment, it is possible to determine whether the vibration of the vibrating object is normal vibration or abnormal vibration using the moving image data packet output from the camera having no moving image compression function.

Embodiments 1 and 2 have been described above. In the first and second embodiments, the vibration detecting device 100 is configured in a functional block form in which each part of the vibration detecting device 100 is independent. However, the configuration of the embodiment described above may not be employed, and the configuration of the vibration detecting device 100 is arbitrary. The functional blocks of the vibration detecting device 100 may be any as long as the functions described in the above embodiments can be realized. These functional blocks may be constituted by any other combination or arbitrary block configuration to constitute the vibration detecting device 100.

Further, the vibration detecting device 100 may not be one device but a system composed of a plurality of devices.

Although the first and second embodiments have been described, the plurality of parts in the embodiments may be combined and implemented. Alternatively, in these embodiments, it is possible to implement one part. In addition, these embodiments may be implemented in whole or in part, in combination.

The embodiments described above are preferably illustrative in nature and are not intended to limit the scope of the invention, the scope of the application, and the scope of the application.

Claims (8)

 A vibration detecting device includes: a normal feature memory unit that memorizes normal feature information indicating a characteristic of normal vibration among vibrations of a vibrating object to be vibrated; and a histogram information generating unit that acquires and displays the aforementioned vibration The moving image data of the moving image of the object extracts an motion vector indicating the motion of the vibrating object from the moving image data, and counts the number of motion vectors for each range of the motion direction range depending on the direction of the direction of the motion vector, and Using the number of motion vectors in each of the foregoing motion direction ranges, a histogram of each data packet of the moving image data is generated as histogram information; and a frequency analysis unit that calculates each using the histogram information a frequency of a time series change of the number of motion vectors in the range of the motion direction and a frequency component of each frequency as vibration characteristic information indicating a characteristic of the vibration of the vibrating object; and a comparing portion from the normal feature memory unit Obtaining the foregoing normal feature information, and comparing the aforementioned vibration feature information with the aforementioned normal feature Information, and determines whether the feature information having the vibration characteristics of normal information.    The vibration detecting device according to the first aspect of the invention, wherein the vibration detecting device includes a normal feature extracting portion that extracts the normal vibration from the vibration characteristic information calculated by the frequency analyzing unit. The characteristic information is used as the aforementioned normal feature information.    The vibration detecting device according to claim 1 or 2, wherein the histogram information generating unit acquires moving image compressed data obtained by compressing the moving image and is a moving image compressed data including the motion vector. The moving image data is extracted from the moving image compressed data.    The vibration detecting device according to claim 1 or 2, wherein the histogram information generating unit acquires a moving image data packet constituting the moving image as the moving image data and the moving image data in proximity The movement of the vibrating object is detected between the packets, thereby extracting the aforementioned motion vector.    The vibration detecting device according to claim 1 or 2, wherein the frequency analyzing unit performs a Fourier transform on a time series change of the number of motion vectors in each of the motion direction ranges, thereby calculating each of the foregoing motion direction ranges. The frequency of the time series change of the number of motion vectors and the frequency component of each frequency serve as the aforementioned vibration characteristic information.    The vibration detecting device according to claim 1 or 2, wherein the vibration detecting device includes a notifying unit that outputs a notification when the comparing portion determines that the vibration characteristic information does not have the normal feature information. Abnormal vibration notification of abnormal vibration.    A vibration detecting method is a vibration detecting method of a vibration detecting device having a normal characteristic memory portion that memorizes normal feature information indicating a characteristic of normal vibration among vibrations of a vibrating object to be vibrated; The vibration detecting method includes the following steps: the histogram information generating unit acquires moving image data indicating a moving image of the vibrating object, and extracts an motion vector indicating the motion of the vibrating object from the moving image data, and belongs to Generating the number of motion vectors for each range of motion directions of the range of motion vectors, and using the number of motion vectors for each of the foregoing motion direction ranges, generating a histogram of each data packet of the aforementioned motion image data as The histogram information; the frequency analysis unit calculates the frequency of the time series change of the number of motion vectors in each of the motion direction ranges and the frequency component of each frequency using the histogram information as the vibration indicating the vibration of the vibrating object. Characteristic vibration characteristic information; and comparison department from the aforementioned normal characteristics The memory unit obtains the normal feature information, compares the vibration feature information with the normal feature information, and determines whether the vibration feature information has the normal feature information.    A vibration detecting program product is a vibration detecting program product having a vibration detecting device of a normal characteristic memory portion, and the normal characteristic memory portion memorizes normal characteristic information indicating a characteristic of normal vibration among vibrations of a vibrating object to be vibrated The vibration detecting program product causes the vibration detecting device belonging to the computer to perform the following processing: a histogram information generating process for acquiring moving image data indicating a moving image of the vibrating object, from the moving image data Extracting an motion vector indicating an action of the vibrating object, counting the number of motion vectors for each range of motion directions in a range of directions of the motion vector, and generating the dynamics by using the number of motion vectors in each of the motion direction ranges The histogram of each data packet of the image data is used as histogram information; the frequency analysis processing uses the histogram information to calculate the frequency and the frequency of the time series change of the number of motion vectors of each of the foregoing motion direction ranges. a frequency component of a frequency to represent the aforementioned vibrating object The vibration characteristic information of the characteristic of the vibration; and the comparison processing, the obtaining the normal feature information from the normal feature memory unit, and comparing the vibration feature information with the normal feature information, and determining whether the vibration feature information has the normal feature information .