Event occurrence period detection system of periodic event

Abstract

The invention provides a system for detecting an event occurrence period of a periodic event. The system comprises a monitoring image capturing device and a period judging device. The monitoring image capturing device is used for periodically capturing a plurality of corresponding static monitoring image data at a plurality of image capturing times. The period judgment device is used for capturing a plurality of feature vectors of the static monitoring image data; calculating a plurality of corresponding vector values from the feature vector to generate a vector value time domain signal; performing a short-time Fourier transform with an analysis window width time, converting the vector value time domain signal to generate a spectrogram, and filtering noise to generate a representative frequency time domain signal; and according to the representative frequency time-domain signal, calculating an event cycle compliance rate and a product term accumulation term number every time the analysis window width time interval is passed, so as to calculate a plurality of event occurrence times for judging the event occurrence cycle.

Description

Technical Field

[0001] The present invention relates to a detection system, and more particularly to a detection system for detecting an event occurrence period of a periodic event. Background Art

[0002] As the industry develops towards automation and intelligence, unmanned and robotized production lines have gradually become the current mainstream development direction. The production and quality monitoring work of the production lines are naturally replaced by automated equipment.

[0003] Since the workpieces or products on the production line are arranged to work periodically according to the established work sequence, once the workpiece or product fails to complete the work that should be completed at one of the workstations of the production line according to the established work cycle, the work of the entire production line will be delayed. Therefore, appropriate devices are needed to monitor the work cycle of each workstation.

[0004] Looking at the existing work cycle detection technologies, most of them rely on a large number of detection devices (such as light blocking end detection devices, touch detection devices, pressure sensing devices or chip detection devices, etc.) to monitor the position of the workpiece or the movement trajectory of the specific mechanism of the working equipment, and then combine the Internet of Things technology with the background big data computing platform to analyze and confirm whether the work that should be completed at each workstation is indeed completed within the allowed work cycle.

[0005] However, since the work content of each workstation is very different, in order to fully realize the above technology, it is necessary to set up a variety of different detection devices to detect different parameters (such as temperature, pressure, number of touches, touch time, light interruption signal interruption time, light interruption signal delay time, light interruption signal duration), etc. according to the actual work content of each workstation. In case of adjustment of the work content of the production line, the type of detection device and the installation location selected also need to be readjusted. In other words, the previous work cycle detection technology generally lacks universality and cannot be used in all workstations. Therefore, it is bound to cost more resources to realize the detection of work event cycles of multiple different workstations. Summary of the invention

[0006] In view of the fact that in the prior art, there is a general lack of universality in event occurrence (work) cycle detection technology, which leads to the problem of needing to spend more resources to realize the event occurrence cycle (work event occurrence cycle of workstation) detection of multiple different periodic events (work events of workstation); the main purpose of the present invention is to provide a new cycle detection technology, so that it can be widely used to detect the event occurrence cycle of multiple different types of periodic events, making it particularly suitable for application in the work cycle detection of multiple different types of workstations.

[0007] One of the necessary technical means adopted by the present invention to solve the problems of the prior art is to provide an event occurrence cycle detection system for a periodic event, which is used to detect an event occurrence cycle of a periodic event and includes a monitoring image capture device and a cycle judgment device.

[0008] The monitoring image capturing device is used to periodically capture a plurality of static monitoring image data corresponding to a monitoring field of view at a plurality of image capturing times according to an image capturing cycle to concatenate into a dynamic monitoring image data, and periodic events occur in the monitoring field of view.

[0009] The period judgment device is communicatively connected to the monitoring image capturing device, is installed with a judgment program, and after executing the judgment program, comprises a feature vector acquisition module, a signal generation module, a spectrum conversion analysis module, a noise filtering module and an event occurrence period judgment module.

[0010] The feature vector extraction module is used to extract dynamic monitoring image data from a monitoring image capture device, and extracts a plurality of feature vectors of a plurality of static monitoring image data corresponding to a plurality of image capture times from the dynamic monitoring image data through a mathematical operation model generated by deep learning.

[0011] The signal generating module is used for calculating the feature vector into a plurality of vector values ​​arranged in the order of occurrence of the image capturing time, and generating a vector value time domain signal according to the image capturing time and the vector value.

[0012] The spectrum conversion analysis module is used for converting the vector numerical time domain signal into a spectrum diagram by performing a short-time Fourier transformation with a wide time interval of an analysis window, and the spectrum diagram includes a plurality of component frequencies and corresponding plurality of component frequency signals and a plurality of component frequency amplitudes.

[0013] The noise filtering module is used to filter out at least one noise signal corresponding to some component frequencies, extract the remaining component frequency signals corresponding to the remaining component frequencies as a plurality of effective component frequency signals, and sequentially extract a component frequency corresponding to one of the effective component frequency signals with the largest amplitude in each of the plurality of analysis window width time intervals as a representative frequency of each analysis window width time interval, thereby generating a representative frequency time domain signal.

[0014] The event occurrence cycle determination module is used to calculate a plurality of event occurrence times according to the representative frequency time domain signal and the analysis window width interval to determine the event occurrence cycle.

[0015] Based on the above necessary technical means, among the derived subsidiary technical means, preferably, the spectrum conversion analysis module is used to define the maximum component frequency amplitude as a maximum amplitude, and define one of the component frequencies corresponding to the maximum amplitude as a main frequency; and the noise filtering module is used to define a noise critical amplitude according to the maximum amplitude, and regard the partial component frequency signals corresponding to the above partial component frequencies corresponding to the component frequency amplitudes less than the noise critical amplitude as at least one noise signal and filter them out, wherein the noise critical amplitude can be 10% of the maximum amplitude. Preferably, the analysis window width time interval is equal to the image acquisition period, and the spectrum conversion analysis module uses the image acquisition period to perform short-time Fourier transform.

[0016] The monitoring image capture device may be an IP Cam, and the periodic event may be a periodic work event of a production line, and the monitoring field of view area is a monitoring field of view area of ​​a workstation of the production line. In addition, the period judgment device may also include a human-machine interface module for a user to set a focus area in the monitoring field of view, and the periodic event occurs periodically in the focus area. Preferably, the human-machine interface module includes a display and an operation interface. The display is used to display dynamic monitoring image data of the monitoring field of view area. The operation interface is used for the user to set the focus area.

[0017] The feature vector extraction module may further include a focus region feature vector extraction unit, and the focus region feature vector extraction unit extracts the feature vector of the static monitoring image data corresponding to the image extraction time from the portion corresponding to the focus region in the dynamic monitoring image data through a mathematical operation model.

[0018] The event occurrence cycle judgment module also includes a compliance rate calculation unit, a statistical unit, an event occurrence time calculation unit and an event occurrence cycle calculation unit. The compliance rate calculation unit accumulates the time interval frequency product terms in sequence to obtain an event cycle compliance rate. The statistical unit is used to count the number of accumulated product terms every time the event compliance rate just reaches an integer, so as to count the corresponding number of accumulated product terms when the event compliance rate just reaches a plurality of the above integers. The event occurrence time calculation unit is used to multiply the number of accumulated product terms by the analytical window width time interval to calculate a plurality of event occurrence times. The event occurrence cycle calculation unit calculates the event occurrence cycle based on at least one event occurrence time difference between any two adjacent event occurrence times.

[0019] In summary, in the event occurrence period detection system of periodic events provided by the present invention, the event occurrence period of periodic events is directly parsed from the static monitoring image data in the dynamic monitoring image data. Therefore, it can be directly used for the detection of the event occurrence periods (such as the work event occurrence periods of the workstation) of multiple different periodic events (such as the periodic work events of the workstation). Therefore, it is no longer necessary to spend more resources to realize the effect of event occurrence period detection of multiple different periodic events.

[0020] Especially when it is applied to workstations, it is no longer necessary to spend more resources to easily detect the work event occurrence cycles of work events in multiple different workstations. Therefore, the execution cost of detecting multiple work event cycles in the production line can be greatly reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 It is a schematic diagram showing a monitoring image capturing device capturing dynamic monitoring image data of a monitoring field of view in a preferred embodiment of the present invention;

[0022] FIG. 2A to FIG. 2E Displays static surveillance image data captured at a portion of the image capture time during an event occurrence cycle;

[0023] Figure 3 A functional block diagram of an event occurrence period detection system for a periodic event provided by a preferred embodiment of the present invention is shown;

[0024] Figure 4 Displayed is a vector value time domain signal waveform diagram generated in a preferred embodiment of the present invention;

[0025] Figure 5 Shown is a spectrum diagram in a preferred embodiment of the present invention;

[0026] Figure 6 Shown is a spectrum diagram after filtering out noise signals in a preferred embodiment of the present invention;

[0027] Figure 7 Shown is a representative frequency-time domain signal waveform diagram in a preferred embodiment of the present invention; and

[0028] Figure 8 The graph shows the corresponding curves of the number of event occurrences and the time of event occurrence of the present invention.

[0029] Component number description

[0030] 100 Detection System

[0031] 200 workstations

[0032] 210 Input conveyor belt

[0033] 220 Work Equipment

[0034] 2201 Robotic Arm

[0035] 2202 Work Platform

[0036] 2203 Detection Probe

[0037] 230 Output conveyor belt

[0038] 300 Circuit Boards to be Tested

[0039] 1 Surveillance image capture device

[0040] 2 Cycle judgment device

[0041] 21 Human Machine Interface Module

[0042] 211 Display

[0043] 212 Operation interface

[0044] JAP Judgment Procedure

[0045] 22 Feature vector extraction module

[0046] 221 Attention region feature vector extraction unit

[0047] 23 Signal generation module

[0048] 231 Vector numerical operation unit

[0049] 232 Signal generation unit

[0050] 24 Spectrum conversion analysis module

[0051] 25 Noise Filter Module

[0052] 251 Noise Filter Unit

[0053] 252 represents frequency signal generating unit

[0054] 26 Event occurrence cycle judgment module

[0055] 261 Compliance Rate Calculation Unit

[0056] 262 Statistics Unit

[0057] 263 Event Occurrence Time Calculation Unit

[0058] 264 Event Occurrence Cycle Operation Unit DETAILED DESCRIPTION

[0059] Since the event occurrence cycle detection system for periodic events provided by the present invention can be widely used to detect the event occurrence cycles of various different periodic events, and is particularly suitable for detecting the event occurrence cycles of periodic work events, its application level is quite broad, so it will not be described one by one here. Only one of the preferred embodiments of its application in a production line workstation is listed for specific explanation, and this embodiment is only used to conveniently and clearly assist in explaining the purpose and efficacy of the embodiment of the present invention.

[0060] See also Figures 1 to 3 , Figure 1 It is a schematic diagram showing a monitoring image capturing device capturing dynamic monitoring image data of a monitoring field of view in a preferred embodiment of the present invention; FIG. 2A to FIG. 2E Displays static surveillance image data captured at a portion of the image capture time during an event cycle. Figures 1 to 3 As shown, a periodic event occurrence cycle detection system (hereinafter referred to as "detection system") 100 is used to detect an event occurrence cycle of a periodic event, and includes a monitoring image capture device 1 and a cycle determination device 2. In this embodiment, the periodic event may be a periodic work event occurring at a workstation 200 of a production line.

[0061] The workstation 200 includes an input conveyor belt 210, a working device 220 and an output conveyor belt 230. The working device 220 includes a robot arm 2201, a working platform 2202 and a detection probe 2203. In this embodiment, the periodic working events occurring in the workstation 200 include the following steps in chronological order: using the input conveyor belt 210 to convey a circuit board 300 to be tested toward the working device 220; using the robot arm 2201 to pick up the circuit board 300 from the input conveyor belt 210; placing the circuit board 300 to be tested on the working platform 2202; extending the detection probe 2203 to contact the circuit board 300 to be tested; retracting the detection probe 2203, and using the robot arm 221 to move the circuit board to be tested from the working platform 222 to the output conveyor belt 230.

[0062] According to the working sequence of the workstation 200, the periodic working events will occur periodically with an event occurrence cycle. The monitoring image capture device 1 can be a network camera (IP Cam) installed in the workstation 200, and according to an image capture cycle, it periodically captures a plurality of static monitoring image data corresponding to a monitoring field of view area at a plurality of image capture times to concatenate into a dynamic monitoring image data, and the periodic events occur in the monitoring field of view area.

[0063] The so-called "image capture cycle" refers to the time interval between any two consecutive captures of static surveillance images, that is, the reciprocal of the frame rate in the dynamic surveillance image data. In this embodiment, the frame rate is 30fps, that is, 30 static surveillance images are captured per second, which is equivalent to an image capture cycle of 1 / 30 second, about 0.033 seconds. In other words, the dynamic surveillance image data is formed by concatenating the static surveillance image data at a frame rate of 30fps.

[0064] The so-called "monitoring field of view" refers to the field of view of the monitoring image capture device 1 that can capture images. In this embodiment, since the periodic event is a periodic work event occurring in the production line 200, the monitoring field of view is a monitoring field of view of a workstation of the production line 200, which covers part of the input conveyor belt 210, the working equipment 200 and part of the output conveyor belt 230, that is, FIG. 2A to FIG. 2E As described above, when the periodic work event proceeds to the instant of the above steps in chronological order, the static monitoring image data captured by the monitoring image capture device 1 includes the following FIG. 2A to FIG. 2E The content displayed.

[0065] The cycle judgment device 2 can be an industrial computer, a desktop computer, a laptop computer, a tablet computer or a smart phone, which is connected to the monitoring image capture device 1 through wired communication or wireless communication means, includes a human-machine interface module 21, and is installed with a judgment program JAP. The human-machine interface module 21 includes a display 211 and an operation interface 212. The display 211 can display dynamic monitoring image data of the monitoring field of view area. The operation interface 212 can be a touch panel, a button or a mouse, etc., for a user (operator of the cycle judgment device 2) to set a region of interest ROI in the monitoring field of view area.

[0066] Generally speaking, the region of interest ROI can be selected from the region where the periodic changes in the dynamic monitoring image data are more significant when the periodic event occurs, or can be selected from the region that can simultaneously cover the locations where multiple different periodic events occur, that is, the region that best reflects the periodic changes of the periodic events. In this embodiment, the region that best reflects the periodic changes of the periodic events is at the work platform 2202, so the region surrounded by the peripheral contour close to the work platform 2202 can be selected as the region of interest ROI.

[0067] After executing the determination program JAP, the period determination device 2 may further include a feature vector extraction module 22 , a signal generation module 23 , a spectrum conversion analysis module 24 , a noise filtering module 25 and an event occurrence period determination module 26 .

[0068] The feature vector extraction module 22 includes a region of interest feature vector extraction unit 221. The feature vector extraction module 22 can directly extract a plurality of feature vectors of static surveillance image data corresponding to the image capture time from the dynamic surveillance image data through a mathematical operation model generated by deep learning. More preferably, the region of interest feature vector extraction unit 221 of the feature vector extraction module 22 can also be used to extract a plurality of feature vectors of static surveillance image data corresponding to the image capture time from the portion of the dynamic surveillance image data corresponding to the region of interest ROI through another mathematical operation model generated by deep learning.

[0069] Regarding the mathematical operation model, the relevant parameters such as chromaticity, saturation, grayscale, sharpness, saturation, etc. of each pixel can be used as the input data for initial learning, and the initial mathematical operation model can be established by using neural network, regression operation or other existing learning algorithms, and then multiple versions of the initial mathematical operation model are modified through multiple deep learning updates. The so-called feature vector is a multi-dimensional vector composed of multiple representative feature variables extracted by the mathematical operation model for each (each frame) static monitoring image data, and each feature variable is regarded as a component.

[0070] For example, the mathematical operation model can extract four representative feature variables A1, B1, C1 and D1 for the first frame of static monitoring image data, and the feature vector corresponding to the first frame of static monitoring image data is (A1, B1, C1, D1). Similarly, the mathematical operation model can extract four corresponding representative feature variables A2, B2, C2 and D2 for the second frame of static monitoring image data, and the feature vector corresponding to the second frame of static monitoring image data is (A2, B2, C2, D2), and so on. In fact, the number of representative feature variables may be several, dozens or even hundreds, depending on the calculation results of the mathematical operation model obtained after multiple learning and modification. The number of feature variables calculated by each different mathematical operation model may also be different.

[0071] It must be emphasized that, since periodic events occur periodically, multiple static surveillance image data may undergo periodic changes. Therefore, no matter which mathematical operation model is used to extract feature vectors from multiple static surveillance image data, these feature vectors will inevitably show periodic changes. Therefore, no matter which mathematical operation model is generated by deep learning, the present invention can be implemented.

[0072] In this embodiment, the image capture period is 1 / 30 second, that is, the image capture time of the first (first frame) static monitoring image data is 1 / 30 second, and its corresponding feature vector is (A1, B1, C1, D1); the image capture time of the second (second frame) static monitoring image data is 2 / 30 second, and its corresponding feature vector is (A2, B2, C2, D2), and so on.

[0073] The signal generation module 23 includes a vector value calculation unit 231 and a signal generation unit 232. The vector value calculation unit 231 is used to calculate the feature vector into a plurality of vector values ​​arranged in the order of occurrence of the image capture time. The vector value can be calculated by the square root method. For example, the vector value V1 of the feature vector (A1, B1, C1, D1) is equal to The vector value V2 of the eigenvector (A2, B2, C2, D2) is equal to

[0074]

[0075] The numerical signal generating unit 232 can generate a vector numerical time domain signal according to the image capture time and the vector numerical value. For example, the first frame time (1 / 30 second) corresponds to the vector numerical value, the second frame time (2 / 30 second) corresponds to the vector numerical value, and so on. Through the corresponding relationship between the vector numerical value and time, a vector numerical time domain signal can be generated, and its waveform diagram is as follows: Figure 4 shown.

[0076] The spectrum conversion analysis module 24 is used to perform a short-time Fourier transform (STFT) on the vector numerical time domain signal with an analysis window width to generate a spectrum diagram (such as Figure 5 As shown), the spectrum diagram includes a plurality of component frequencies and corresponding plurality of component frequency signals and a plurality of component frequency amplitudes. Short-time Fourier Transform (STFT) is to cut the vector numerical time domain signal into a plurality of local signals with a time interval equal to the resolution window width, and then perform Fourier transform on the plurality of local signals. Therefore, the resolution window width should be an integer multiple of the image acquisition period (1 / 30 second), preferably 1, that is, the resolution window width is exactly equal to the image acquisition period.

[0077] The spectrum conversion analysis module 24 can define the largest one of the component frequency amplitudes as a maximum amplitude for each analysis window width time interval, and define one of the component frequencies corresponding to the maximum amplitude as a main frequency. In this embodiment, the first analysis window width time interval is 0 to 1 / 30 seconds, and the second analysis window width time interval is 1 / 30 to 2 / 30 seconds. In this embodiment, the main frequency is approximately 1 / 70 Hz (approximately equal to 0.014 Hz).

[0078] The noise filtering module 25 includes a noise filtering unit 251 and a representative frequency signal generating unit 252. The noise filtering unit 251 can define a noise critical amplitude according to the maximum amplitude, and regard the component frequency signals corresponding to the component frequency amplitudes smaller than the noise critical amplitude as the at least one noise signal and filter them out. Preferably, the noise critical amplitude can be 10% of the maximum amplitude. After the noise is filtered out by the noise filtering unit 251, Figure 5 The spectrum shown will be transformed into Figure 6 The spectrum diagram shown.

[0079] The representative frequency signal generating unit 252 can extract the remaining component frequency signals corresponding to the remaining component frequencies as a plurality of effective component frequency signals, and sequentially extract the component frequency corresponding to one of the effective component frequency signals with the largest amplitude in each analysis window width time interval as a representative frequency of each analysis window width time interval, thereby generating a representative frequency time domain signal. That is, the frequency with the largest amplitude in the first analysis window width time interval (0 to 1 / 30 seconds) is extracted as the representative frequency of the first analysis window width time interval (0 to 1 / 30 seconds), and the frequency with the largest amplitude in the second analysis window width time interval (1 / 30 to 2 / 30 seconds) is extracted as the representative frequency of the second analysis window width time interval (0 to 1 / 30 seconds), and so on, and the following can be obtained. Figure 7 The representative frequency time domain signal waveform is shown in the figure. Because each analysis window width corresponds to a representative frequency, the representative frequency time domain signal waveform will be a square wave with amplitude variation. In this embodiment, the main frequency is about 1 / 70 Hz (about 0.014 Hz). Figure 7 It can be seen that the representative frequency corresponding to most of the analysis window widths is 1 / 70 Hz (approximately equal to 0.014 Hz).

[0080] The event occurrence period determination module 26 may include a standard achievement rate calculation unit 261 , a statistical unit 262 , an event occurrence time calculation unit 263 and an event occurrence period calculation unit 264 .

[0081] The achievement rate calculation unit 261 is used to multiply each analysis window width time interval with the corresponding representative frequency to calculate a time interval frequency product term, and sequentially accumulate the time interval frequency product terms to obtain an event cycle achievement rate. Specifically, the event cycle achievement rate (ECAR) can be calculated by the following formula:

[0082]

[0083] in:

[0084] i represents the width and time interval of the i-th analysis window;

[0085] f i Indicates the representative frequency corresponding to the width and time interval of the i-th analysis window, in Hz;

[0086] T W represents the time interval of the analysis window width, which is 1 / 30 second in this embodiment;

[0087] n represents the number of accumulated product terms of time-frequency product terms.

[0088] The event cycle achievement rate ECAR equal to 1 means that one periodic event is completed, the event cycle achievement rate ECAR equal to 2 means that two periodic events are completed, and so on.

[0089] The statistical unit 262 is used to count a product term accumulation number every time the event achievement rate just reaches an integer, and to count the corresponding plural product term accumulation numbers when the event achievement rate just reaches a plurality of the above integers. In this embodiment, when the product term accumulation number n is approximately equal to 2100, 4201, 6300..., the event period achievement rate ECAR just reaches 1, 2, 3... respectively. The so-called "just reaches" means "just equal to" or "if there is no exactly equal, then the one that is greater than but closest is considered just reached".

[0090] For example: if the number of accumulated items n is 2100, the event cycle compliance rate ECAR=1, then the number of accumulated items n of the product items when the event cycle compliance rate ECAR just reaches 1 is 2100. Another example: if the number of accumulated items n of the product items is 2099, the event cycle compliance rate ECAR=0.9998; the number of accumulated items n of the product items is 2100, the event cycle compliance rate ECAR=1.0001; the number of accumulated items n of the product items is 2101, the event cycle compliance rate ECAR=1.0003, then the number of accumulated items n when the event cycle compliance rate ECAR just reaches 1 is 2100.

[0091] The event occurrence time calculation unit 263 is used to multiply the number of product term accumulation items by the analysis window width time interval to calculate the multiple event occurrence times. In this embodiment, when the event cycle standard rate ECAR=1, the number of product term accumulation items n (2100) multiplied by the analysis window width time interval (1 / 30) can calculate the event occurrence time of the first periodic event as 70 seconds; when the event cycle standard rate ECAR=2, the number of product term accumulation items n (4201) multiplied by the analysis window width time interval (1 / 30) can calculate the event occurrence time of the second periodic event as 140.03 seconds; when the event cycle standard rate ECAR=3, the number of product term accumulation items n (6300) multiplied by the analysis window width time interval (1 / 30) can calculate the event occurrence time of the third periodic event as 210 seconds; and so on.

[0092] Since the event period achievement rate ECAR reflects the number of occurrences of periodic events, it can also be replaced by the number of event occurrences, thus generating Figure 8 The corresponding curve diagram of the number of event occurrences and the time of event occurrence is shown. Among them, each dot is a data point where the number of event occurrences is an integer. From left to right, the first dot represents the time of the first periodic event, the second dot represents the time of the second periodic event, and so on.

[0093] The event occurrence cycle calculation unit 264 can calculate an event occurrence time difference according to the time difference between two adjacent event occurrence times. Preferably, the event occurrence time differences can be averaged to form the event occurrence cycle.

[0094] In this embodiment, the event occurrence time of the first periodic event is 70 seconds, the event occurrence time of the second periodic event is 140.03 seconds, and the event occurrence time of the third periodic event is 210 seconds. Therefore, the event occurrence time difference between the second periodic event and the first periodic event is 70.03 seconds, and the event occurrence time difference between the third periodic event and the second periodic event is 69.97 seconds. After the arithmetic average of the two, the event occurrence cycle can be obtained as 70 seconds. Alternatively, the remaining multiple event occurrence time differences can be averaged with the above two event occurrence time differences to calculate a more accurate event occurrence cycle.

[0095] In practice, in addition to using the above method to determine a more accurate event occurrence period, the event occurrence period determination module 26 can also directly find the most representative frequency corresponding to all analysis window width intervals in the representative frequency time domain signal waveform diagram, and use it as the main frequency. In this embodiment, the most representative frequency is about 0.014 Hz (equivalent to about 1 / 70 Hz), indicating that the main frequency is about 1 / 70 Hz, and then the inverse of the main frequency is directly taken to roughly calculate that the event occurrence period is about 70 seconds.

[0096] At the same time, Figure 8 In the figure, the 30th dot is at about 2100 seconds, which means that 30 periodic events can be completed in about 2100 seconds. It can also be verified that the event cycle is about 70 seconds. Figure 7 and Figure 8 It can also be seen that between about 2100 seconds and 2600 seconds, the representative frequencies of each analysis window width are all zero, and there are no dot marks, indicating that no periodic events occur during the period from about 2100 seconds to 2600 seconds.

[0097] Since the above-mentioned feature vector extraction module 22, signal generation module 23, spectrum conversion analysis module 24, noise filtering module 25 and event occurrence period judgment module 26 are all generated by executing the judgment program JAP, the feature extraction module 21, supervised learning module 22, judgment module 23 and verification warning module 24 can essentially be the (partial) main program, subroutine of the judgment program JAP or the program page or function interface generated after executing the judgment program JAP. Anyone with ordinary knowledge in the relevant technical field (especially the field of artificial intelligence algorithms) can use appropriate programming languages ​​to write the judgment program JAP (including its main program or subroutine) with the functions of the above-mentioned feature vector extraction module 22, signal generation module 23, spectrum conversion analysis module 24, noise filtering module 25 and event occurrence period judgment module 26 according to the above learning and judgment logic, so as to realize the above-mentioned various technologies of the present invention.

[0098] In summary, in the event occurrence period detection system 100 of periodic events provided by the present invention, the event occurrence period of periodic events is directly parsed from the static monitoring image data in the dynamic monitoring image data. Therefore, it can be directly used for the detection of the event occurrence period (such as the work event occurrence period of the workstation 200) of multiple different periodic events (such as the periodic work events of the workstation 200), thereby achieving the effect of realizing the event occurrence period detection of multiple different periodic events without spending more resources.

[0099] Especially when it is applied to workstations, it is no longer necessary to spend more resources to easily detect the work event cycles of multiple different workstations. Therefore, the execution cost of detecting multiple work event cycles in the production line can be greatly reduced.

[0100] The purpose of the detailed description of the preferred embodiments above is to more clearly describe the features and spirit of the present invention, rather than to limit the scope of the present invention by the preferred embodiments disclosed above. On the contrary, the purpose is to cover various changes and arrangements with equivalents within the scope of the patent scope to be applied for by the present invention.

Claims

1. A system for detecting the occurrence period of a periodic event, characterized in that: An event occurrence period for detecting a periodic event, and comprising: A surveillance image capturing device, used for periodically capturing a plurality of static surveillance image data corresponding to a surveillance field of view at a plurality of image capturing times according to an image capturing cycle, so as to be serially connected into a dynamic surveillance image data, wherein the periodic event occurs in the surveillance field of view; and A period determination device is communicatively connected to the monitoring image capturing device, is installed with a determination program, and after executing the determination program, comprises: a feature vector extraction module, for extracting the dynamic monitoring image data from the monitoring image capture device, and extracting a plurality of feature vectors of the plurality of static monitoring image data corresponding to the plurality of image capture times from the dynamic monitoring image data through a mathematical operation model generated by deep learning; a signal generating module, used for calculating the plurality of feature vectors into a plurality of vector values ​​arranged in the order of occurrence of the plurality of image capture times, and generating a vector value time domain signal according to the image capture time and the plurality of vector values; A spectrum conversion analysis module, used for converting the vector numerical time domain signal into a spectrum diagram by performing a short-time Fourier transform with a wide time interval of an analysis window, and the spectrum diagram includes a plurality of component frequencies and corresponding plurality of component frequency signals and a plurality of component frequency amplitudes; a noise filtering module, for filtering out at least one noise signal corresponding to some of the component frequencies, extracting the remaining component frequency signals corresponding to the remaining component frequencies as a plurality of effective component frequency signals, and sequentially extracting a component frequency corresponding to one of the effective component frequency signals with the largest amplitude in each of the plurality of analysis window widths and time intervals as a representative frequency of each of the analysis window widths and time intervals, thereby generating a representative frequency time domain signal; and An event occurrence period determination module is used to calculate a plurality of event occurrence times according to the representative frequency time domain signal and the analysis window width interval to determine the event occurrence period.

2. The event occurrence period detection system of a periodic event according to claim 1, characterized in that: The spectrum conversion analysis module is used to define the largest amplitude of the component frequencies as a maximum amplitude, and define one of the component frequencies corresponding to the maximum amplitude as a main frequency; and the noise filtering module is used to define a noise critical amplitude based on the maximum amplitude, and regard the part of the component frequency signals corresponding to the above-mentioned part of the component frequencies corresponding to the component frequency amplitudes smaller than the noise critical amplitude as the at least one noise signal and filter them out.

3. The event occurrence period detection system of a periodic event according to claim 2, characterized in that: The noise critical amplitude is 10% of the maximum amplitude, and the noise filtering module is used to regard the portion of the component frequency signals corresponding to the portion of the component frequencies corresponding to the component frequency amplitudes less than 10% of the maximum amplitude as the at least one noise signal and filter them out.

4. The event occurrence period detection system of a periodic event according to claim 1, characterized in that: The analysis window width is equal to the image acquisition period, and the spectrum conversion analysis module is used to perform the short-time Fourier transform with the image acquisition period.

5. The event occurrence period detection system of a periodic event according to claim 1, characterized in that: The period determination device further includes a human-machine interface module, wherein the human-machine interface module is used for a user to set a focus area in the monitoring field of view, and the periodic event occurs periodically in the focus area.

6. The event occurrence period detection system of a periodic event according to claim 5, characterized in that: The human-machine interface module comprises: A display, for displaying the dynamic monitoring image data of the monitoring field of view; and An operation interface is used for the user to set the area of ​​interest.

7. The event occurrence period detection system of a periodic event according to claim 1, characterized in that: The feature vector capture module also includes a focus area feature vector capture unit, and the focus area feature vector capture unit is used to capture the multiple feature vectors of the multiple static monitoring image data corresponding to the multiple image capture times from the part of the dynamic monitoring image data corresponding to the focus area through the mathematical operation model.

8. The event occurrence period detection system of a periodic event according to claim 1, characterized in that: The event occurrence period determination module further comprises: a compliance rate calculation unit, used for multiplying each of the analytical time intervals by the corresponding representative frequency to calculate a time interval frequency product term, and sequentially accumulating the time interval frequency product terms to obtain an event cycle compliance rate; a statistical unit, configured to calculate a number of cumulative product terms each time the event target achievement rate reaches an integer, and to calculate a corresponding number of cumulative product terms each time the event target achievement rate reaches a plurality of integers; an event occurrence time calculation unit, used for multiplying the plurality of product terms and the number of accumulated terms by the analysis window width time interval respectively, so as to calculate a plurality of event occurrence times; An event occurrence period calculation unit is used to calculate the event occurrence period according to at least one event occurrence time difference between any two adjacent event occurrence times.

Images