JP2009210490A - Image analysis system and computer program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To obtain output results of high reliability by enabling PIV analysis due to an image correlation method using light and shade images of measuring objects. <P>SOLUTION: An image processing means 43 includes a filtering means 431 for composing a high pass filter cutting a spatial frequency constituting in a prescribed frequency or lower from the image data of two consecutive times. As a result, the peak of a correlation coefficient is clear, and a moving distance between prescribed light and shade images in the image data of the two times and a moving direction (speed vector) are obtained by applying PIV analysis. When it is determined that an error vector is contained in the obtained speed vector, a cut-off frequency is regulated again to perform filtering due to a filtering means 431, and PIV analysis is performed again, concerning the image data processed by filtering. As a result, the speed vector which does not contain the error vector is obtained. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to an image analysis system for analyzing a velocity field of a measurement target and a computer program used for the image analysis system.

As a system for observing smoke, water vapor, volcanic ash, yellow sand, and the like discharged from a chimney of a power generation facility or the like using a particle image velocimetry (hereinafter referred to as “PIV”), the present applicant has disclosed Patent Document 1 Proposed the technology disclosed in. This enables measurement from a long distance by extracting the turbulent flow structure of the fluid to be measured. By considering the turbulent flow structure as a pseudo particle image, the PIV method is applied to obtain fluid velocity data. When extracting the turbulent flow structure, the spatial frequency is obtained by using the Strouhal number and the representative length of the object that generates the turbulent flow structure of the fluid to be measured, and extracted by applying a high-pass filter process at this spatial frequency. is doing.
WO2005 / 095993A1 publication

  In Patent Document 1, PIV analysis, in particular, PIV analysis using an image correlation method, can be applied to a measurement object separated by a long distance by regarding the turbulent flow structure as a pseudo particle image. However, the technique of Patent Document 1 can measure only those that can extract a turbulent structure. For example, it is difficult to measure a non-fluid or laminar fluid. Therefore, it is conceivable to use the grayscale images of these measurement objects in the analysis including the velocity field of laminar fluid and non-fluid velocity field (behavior change due to vibration) without extracting the turbulent flow structure. Conventionally, however, in the case of a grayscale image, it has been difficult to use the image correlation method because the peak of the correlation coefficient is unclear or multiple peaks appear. In view of this point, the present applicant, as Japanese Patent Application No. 2006-325235, creates a pseudo particle image by clarifying the peak value of the correlation coefficient by applying a high-pass filter to image data consisting of a grayscale image, thereby It is proposed to analyze the velocity field of grayscale image using correlation method. However, it has been found that when the high-pass filter is applied, if the set cutoff frequency is not appropriate, the reliability of the analysis result by PIV is affected.

  The present invention has been made in view of the above, and enables PIV analysis by an image correlation method using a gray-scale image to be measured, and analyzes not only turbulent flow but also laminar fluid and non-fluid velocity fields. It is an object of the present invention to provide an image analysis system capable of performing an analysis result with high reliability and a computer program used for the image analysis system.

  In order to solve the above-described problem, the image analysis system of the present invention compares time-series image data obtained by imaging a measurement target with an imaging unit at a predetermined imaging time interval, and compares the time series image data with the predetermined data included in the image data. An image analysis system comprising image processing means for analyzing the velocity field of a grayscale image, wherein the image processing means cuts a spatial frequency component below a predetermined frequency from image data at two or more consecutive times A spatial frequency component after filtering by the filtering means, and converting the velocity vector between predetermined grayscale images included in the obtained image data at two or more times by PIV analysis. A PIV analysis means to be obtained, and an error vector is included in the velocity vector of the grayscale image obtained by the PIV analysis means. When it is determined to be, the filtering means adjusts the cutoff frequency filtering the image data again, characterized in that the PIV analysis is again performed by the PIV analysis means.

  In addition, the image analysis system of the present invention compares time-series image data obtained by imaging the measurement object by the imaging means at a predetermined imaging time interval, and compares the speed of the predetermined grayscale image included in the image data. An image analysis system including an image processing unit for analyzing a field, wherein the image processing unit includes a high-pass filter that cuts a spatial frequency component of a predetermined frequency or less from image data at two or more consecutive times. And PIV analysis means for converting a spatial frequency component after filtering by the filtering means into an image and obtaining a velocity vector between predetermined grayscale images included in the obtained image data at two or more times by PIV analysis, It is determined whether or not an error vector is included in the velocity vector of the grayscale image obtained by the PIV analysis means. When the determination unit determines that the erroneous vector is included by the determination unit, the cutoff frequency of the filtering unit is adjusted, and the filtering unit performs a process of filtering the image data again. It is preferable that the apparatus includes a reprocessing command unit that re-executes the PIV analysis by the PIV analysis unit.

  The determination unit of the image analysis system includes a unit that determines, from among the plurality of velocity vectors obtained by the PIV analysis unit, a velocity vector that shows a relatively different tendency with respect to the magnitude or direction of the velocity vector as an erroneous vector. It is preferable.

  The filtering unit of the image analysis system is set to perform a plurality of filtering processes at different cutoff frequencies for one image data, and the PIV analysis unit is configured to perform image filtering between the image data having the same cutoff frequency by the filtering unit. A speed vector is obtained by PIV analysis, and the determination means compares a plurality of speed vectors obtained for each cut-off frequency, and determines whether the magnitude of the speed vector or the degree of variation in the direction exceeds a predetermined threshold value. It is preferable to have a configuration including means for determining.

  Note that it is preferable that the image pickup unit of the image analysis system is a long-distance type that includes a long-focus optical system and can image the measurement object separated by a long distance.

  Further, the computer program of the present invention compares time-series image data obtained by imaging the measurement object by the imaging means at a predetermined imaging time interval, and compares the speed field of the predetermined grayscale image included in the image data. A computer program that constitutes an image processing means in an image analysis system for analyzing the image, the filtering means comprising a high-pass filter that cuts a spatial frequency component below a predetermined frequency from image data at two or more consecutive times, and the filtering PIV analysis means for converting a spatial frequency component after being filtered by the means into an image and obtaining a velocity vector between predetermined grayscale images included in the obtained image data at two or more times by PIV analysis, and the PIV analysis An error vector is included in the velocity vector of the grayscale image obtained by the means A determination unit that determines whether or not an error vector is included by the determination unit, and adjusts a cutoff frequency of the filtering unit and filters the image data again by the filtering unit. And a reprocessing command means for re-executing the PIV analysis by the PIV analyzing means.

  The determination means of the computer program includes means for determining, as an erroneous vector, a speed vector that shows a relatively different tendency with respect to the magnitude or direction of the speed vector among the plurality of speed vectors obtained by the PIV analysis means. Is preferred.

  The filtering means of the computer program is set so as to perform a plurality of filtering processes at different cutoff frequencies for one image data, and the PIV analyzing means performs PIV between image data of the same cutoff frequency by the filtering means. A speed vector is obtained by analysis, and the determination means compares a plurality of speed vectors obtained for each cut-off frequency, and determines whether the magnitude of the speed vector or the degree of variation in the direction exceeds a predetermined threshold value. It is preferable to have a means to do.

  According to the present invention, the image processing means includes filtering means that constitutes a high-pass filter that cuts a spatial frequency component equal to or lower than a predetermined frequency from image data at two or more consecutive times. As a result, the peak of the correlation coefficient is clarified, and PIV analysis is applied to obtain the moving distance and moving direction between predetermined grayscale images in the image data at two or more times. Therefore, the measurement target is not limited to turbulent flow, and laminar fluid and non-fluid velocity fields can be analyzed. In addition, in the present invention, when it is determined that an error vector is included in the obtained velocity vector, the cutoff frequency is adjusted, and filtering by the filtering unit is performed again. This is a configuration in which PIV analysis is performed again. As a result, a velocity vector that does not include an erroneous vector is obtained, and a highly reliable velocity field analysis can be performed.

  Hereinafter, embodiments of the present invention will be described in more detail with reference to the drawings. FIG. 1 shows an image analysis system 1 according to an embodiment of the present invention, which includes a CCD camera 2 having a long focus optical system 3, a computer 4 and the like as imaging means.

  A long-focus optical system 3 is attached to the CCD camera 2, and it is preferable to use a single-focus lens (hereinafter referred to as “single lens”) as the long-focus optical system 3. In this case, it is more preferable to provide a turret so that a plurality of types of single lenses can be selected. By using the turret, a single lens can be automatically selected. In the case of a lens having a zoom function, it is generally disadvantageous that the curvature of field is large. However, any lens that can provide a stable image with a high refractive index glass can be used. In this embodiment, a camera (CCD camera) provided with a CCD image pickup device is used as the image pickup means, but a camera provided with a CMOS image pickup device may be used instead.

  As shown in FIGS. 1 and 2, the computer 4 is connected to the CCD camera 2 and receives a control means 41 for controlling the driving of the CCD camera 2 and an image signal photographed by the CCD camera 2 to receive a predetermined value. An image capturing means 42 and an image processing means 43 for performing processing are provided. The CCD camera 2 continuously captures the grayscale image to be measured at minute time intervals. Since the system analyzes the grayscale image of the measurement target, the measurement target includes not only turbulent and laminar fluids such as smoke, clouds, volcanic ash, and yellow sand, but also non-fluids such as structures such as buildings, steel towers, and chimneys. It is. In the case of a non-fluid, a velocity field such as vibration of the structure can be analyzed. Further, since a grayscale image is captured, it is not necessary to spray tracer particles or the like in a measurement object such as a fluid or special marking or the like in a non-fluid measurement object. The control means 41 adjusts the focal length of the CCD camera 2 and the like. The image capturing means 42 includes a frame grabber board for digitizing the grayscale image signal from the CCD camera 2. The image processing means 43 is composed of a computer program, and analyzes the digital image signal output from the frame grabber board by the image correlation method. Note that a circuit for correcting distortion aberration of an image or the like may be provided in the preceding stage of the image processing means 43.

  In the image processing means 43, a predetermined grayscale image in two-time image data captured by the CCD camera 2 at a minute time interval is compared and analyzed by an image correlation method, and a velocity field to be measured is analyzed. Specifically, between two acquired images (first time image and second time image), the search area at the second time is shifted by a predetermined pixel with respect to the reference area at the first time, The position of the search area is specified by comparing the similarity of the light and shade distribution information. The density distribution information is expressed as a spatial frequency component, and specifies the position of the search region where the correlation coefficient of each spatial frequency component in the reference region and the search region is maximum. A velocity vector is obtained from the positional relationship between the specified search area and reference area.

  More specifically, as shown in FIG. 3, the image processing unit 43 includes a filtering unit 431, a PIV analysis unit 432, a determination unit 433, and a reprocessing command unit 434, which are computer programs. When trying to find the velocity vector to be measured from the spatial frequency components of the grayscale image, if all the spatial frequency components are analyzed, the peak value of the correlation coefficient is unclear or there are multiple peak values. Since the spatial frequency component exists and it is difficult to specify the true peak value, the filtering unit 431 performs filtering that leaves only the necessary spatial frequency component as the preprocessing unit (preprocessing step) of the PIV analysis unit 432. The filtering unit 431 according to the present embodiment includes a high-pass filter that cuts a spatial frequency component below a predetermined frequency. Although the value of the cut-off frequency in the high-pass filter is arbitrary, in this embodiment, the cut-off frequency can be adjusted.

  The PIV analysis unit 432 is set with a computer program for image correlation that performs PIV analysis. The image correlation method includes a direct cross-correlation method, an FFT cross-correlation method, an autocorrelation method, a laser speckle method, a luminance difference accumulation method, and the like, and any of these can be applied. However, it is effective when applied to the direct cross correlation method and the FFT cross correlation method.

  The determination unit 433 determines whether or not an error vector is included in the velocity vector obtained by the PIV analysis unit 432. For example, assuming that the flow speed and direction are relatively stable, a velocity vector as shown in FIG. 4 (a) should be output, but when a velocity vector as shown in FIG. 4 (b) is output. Is considered to contain an erroneous vector. As a method for determining whether or not an erroneous vector is included in the output of FIG. 4B, a speed vector that shows a relatively different tendency with respect to the magnitude or direction of the speed vector among the plurality of speed vectors. Can be considered as a false vector. Whether or not they show relatively different tendencies, for example, when each velocity vector deviates from a statistically determined range of variance with respect to the magnitude or direction of the velocity vector, It can be determined that it is shown.

  When the determination unit 433 determines that an erroneous vector is included, the reprocessing command unit 434 adjusts the cutoff frequency of the high-pass filter that constitutes the filtering unit 431 and executes the filtering process and the PIV analysis process again. It is a means for instructing the filtering means 431 and the PIV analysis means 432 so as to make them. By adjusting the cutoff frequency of the high-pass filter, the spatial frequency component to be filtered changes, and the filtered image data becomes different. For example, if the cutoff frequency is raised and filtered, the image data is extracted with a finer pattern, and if the cutoff frequency is lowered and filtered, the image data is left with a gentler pattern. The PIV analysis is performed again on the image data that has been filtered again in this way.

  Next, the image analysis process of the present embodiment will be described based on FIG. First, the CCD camera 2 captures an image of the measurement target at a time interval Δt, and N time-series grayscale images are acquired (S101). The acquired grayscale image is input as a digital image signal to the image processing means 43 through the image capturing means 42. The input digital image signal is filtered by the high-pass filter of the filtering unit 431 for all N images or two consecutive images in time series, with a spatial frequency component equal to or lower than a preset cutoff frequency. (S102).

  Next, the PIV analysis unit 432 performs PIV analysis using the image correlation method (S103), and the determination unit 433 determines whether or not an error vector is included in the obtained velocity vector as described above. (S104). If it is determined that no erroneous vector is included, the result of the PIV analysis output in S103 is adopted. If it is determined that an erroneous vector is included, the reprocessing instruction unit 434 instructs the filtering unit 431 and the PIV analysis unit 432 to re-execute the filtering process and the PIV analysis process (S105). When the re-execution is instructed, the process returns to S102. Returning to S102, the filtering unit 431 performs the filtering process by resetting the cutoff frequency to a value different from the previous one, and the PIV analysis process (S103) is performed on the image data subjected to the filtering process again. As a result, the obtained speed vector is again determined by the determining means 433, and when it is determined that the erroneous vector is not included, the speed vector obtained this time is adopted as the analysis result and includes the erroneous vector. If it is determined that the frequency is determined, filtering is performed by resetting the cutoff frequency again in accordance with a command from the reprocessing command unit 434 (S105). Thereafter, these steps are repeated, and a velocity vector not including an erroneous vector is obtained.

  FIG. 6 is a diagram showing an image analysis process of an image analysis system according to another embodiment of the present invention. In this embodiment, the CCD camera 2 captures a measurement target at a time interval Δt and obtains N time-series grayscale images (S201) in the same manner as in the above embodiment, but the high-pass filter of the filtering unit 431 is used. In the filtering process according to, a plurality of cutoff frequencies are used, and the same image is filtered at different cutoff frequencies (S202). For example, three “normal value”, “normal value + 10% value”, and “normal value−10% value” are prepared as the cut-off frequencies, and image data filtered with these three cut-off frequencies is obtained. That is, in this case, three pieces of image data subjected to filtering processing are obtained for one grayscale image captured by the CCD camera 2.

  Next, PIV analysis using each image data is performed (S203). The PIV analysis is performed between image data having the same cutoff frequency, that is, between image data of “normal value”, image data of “normal value + 10% value”, and image data of “normal value−10% value”. . As a result, three types of velocity vector displays are obtained as the PIV analysis result, and the determination unit 433 compares the three types of velocity vector displays related to the same grayscale image with each other, and whether they show the same tendency. It is determined whether or not (S204). Whether or not the same tendency is indicated is determined by comparing a plurality of obtained velocity vectors and determining whether the degree of relative variation in the magnitude or direction of the velocity vector statistically exceeds a predetermined threshold. . As a result, when the variations of the three types of speed vector display are all within the predetermined threshold, it is assumed that the PIV analysis result is correct, and one of the output results, for example, the output result of “normal value” is adopted. To do. If only one of the three types deviates from the predetermined threshold and the remaining two variations are within the predetermined threshold, the remaining two PIV analysis results are assumed to be correct, and one of the output results is selected. adopt.

  On the other hand, if all three types of variations do not fall within the predetermined threshold, the reprocessing command unit 434 commands re-execution of the filtering process and the PIV analysis process (S205). When the re-execution is instructed, the filtering unit 431 sets a value different from the previous value as the “normal value” and, for example, sets the “+ 10% value” and “−10% value” again to perform the filtering process. Do. Thereafter, these steps are repeated to obtain a velocity vector that does not include an erroneous vector. Note that when only one of the three types falls outside the predetermined threshold and the remaining two variations are within the predetermined threshold, the remaining two PIV analysis results are not treated as correct, and in that case, the cut It is also possible to set so that the off-frequency is reset and the filtering process is performed again. In the present embodiment, three types of cut-off frequencies are set, but it is of course not limited to three types, and may be two types or four or more types.

  In each of the above-described embodiments, the determination by the determination unit 433 is automatically performed. From the result of the PIV analysis, the measurer determines whether to reset the cutoff frequency and perform filtering again. It is also possible. For example, when the flow speed and direction are relatively stable, the rule of thumb is that a velocity vector as shown in FIG. When a speed vector is output, it is possible to determine that an erroneous vector is included therein and reset the cutoff frequency.

  In each of the above-described embodiments, a high-pass filter is provided as the filtering unit 431 and only low frequency components below a predetermined cutoff frequency are cut. However, when a plurality of correlation coefficient peaks appear. In addition, as proposed in Japanese Patent Application No. 2006-325235, it is also possible to provide a low-pass filter that cuts a spatial frequency component of high frequency above a predetermined level.

  In the present invention, it is possible to analyze the behavior of the fluid to be measured or non-fluid from the grayscale image, and it is not necessary to spray tracer particles or the like, and no special marking or the like is required on the structure. Therefore, it is possible to analyze a flow field of a distant fluid that is difficult to access and a non-fluid movement (building vibration, etc.) by imaging using a long focal point optical system. Thus, for example, power plant operation management by analyzing chimney smoke flow fields, operation management by analyzing steam flow fields from cooling towers of nuclear power plants and geothermal power plants, volcanic ash and yellow sand flow fields It can be used for environmental impact assessment by analyzing In addition, it is possible to analyze the flow field of smoke generated from a large-scale fire site and contribute to countermeasures and evacuation guidance. In addition, by analyzing the flow field of the cloud (cloud bottom), it can be used for regional weather forecasts, analysis of the wind around the transmission line and transmission tower, measurement of pollen flow, structure of the tower, etc. It can also be used for vibration analysis of objects. In the case of volcanic eruptions and large-scale fires, the image analysis system of the present invention can be installed in a vehicle and the behavior of the measurement target can be analyzed while moving. It is also useful for effective disaster countermeasures. Note that the distance from the long focus optical system to the fluid to be measured varies depending on the accuracy of the long focus optical system and the imaging device used, and is not particularly limited. In view of the above, it is practically preferable to use at 10 to 20 km.

FIG. 1 is a diagram showing an overview of an image analysis system according to an embodiment of the present invention. FIG. 2 is a block diagram illustrating a schematic configuration of a computer of the image analysis system according to the embodiment. FIG. 3 is a block diagram showing a configuration of image processing means set in the computer according to the embodiment. FIG. 4A is a diagram schematically illustrating an example of a velocity vector display, and FIG. 4B is a diagram schematically illustrating an example of a velocity vector display including an erroneous vector. FIG. 5 is a diagram for explaining an image processing process by the image analysis system according to the embodiment. FIG. 6 is a diagram for explaining an image processing process by an image analysis system according to another embodiment of the present invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Image analysis system 2 CCD camera 3 Long focus optical system 4 Computer 41 Control means 42 Image capture means 43 Image processing means 431 Filtering means 432 PIV analysis means 433 Determination means 434 Reprocessing command means

Claims (8)

Image processing means for comparing the time-series image data obtained by imaging the measurement object by the imaging means at predetermined imaging time intervals and analyzing the velocity field of the predetermined grayscale image included in the image data is provided. An image analysis system,
The image processing means includes
Filtering means comprising a high-pass filter for cutting spatial frequency components below a predetermined frequency from image data at two or more consecutive times;
PIV analysis means for converting a spatial frequency component after filtering by the filtering means into an image and obtaining a velocity vector between predetermined grayscale images included in the obtained image data at two or more times by PIV analysis. ,
When it is determined that an error vector is included in the velocity vector of the grayscale image obtained by the PIV analysis unit, the filtering unit adjusts a cutoff frequency and filters the image data again, An image analysis system characterized in that PIV analysis by the PIV analysis means is re-executed. Image processing means for comparing the time-series image data obtained by imaging the measurement object by the imaging means at predetermined imaging time intervals and analyzing the velocity field of the predetermined grayscale image included in the image data is provided. An image analysis system,
The image processing means includes
Filtering means comprising a high-pass filter for cutting spatial frequency components below a predetermined frequency from image data at two or more consecutive times;
PIV analysis means for converting a spatial frequency component after filtering by the filtering means into an image and obtaining a velocity vector between predetermined grayscale images included in the obtained image data at two or more times by PIV analysis;
Determination means for determining whether an error vector is included in the velocity vector of the grayscale image obtained by the PIV analysis means;
When the determination unit determines that an erroneous vector is included, the cutoff frequency of the filtering unit is adjusted, and the filtering unit executes the process of filtering the image data again, and the PIV analysis unit An image analysis system comprising reprocessing command means for re-executing PIV analysis.   The determination means includes means for determining, from among the plurality of speed vectors obtained by the PIV analysis means, a speed vector showing a relatively different tendency with respect to the magnitude or direction of the speed vector as an erroneous vector. The image analysis system according to claim 2. The filtering means is set to perform a plurality of filtering processes at different cutoff frequencies for one image data,
The PIV analysis means obtains a velocity vector by PIV analysis between image data of the same cutoff frequency by the filtering means,
The determination means comprises means for comparing a plurality of velocity vectors obtained for each cutoff frequency and determining whether the magnitude of the velocity vector or the degree of variation in direction exceeds a predetermined threshold value. The image analysis system according to claim 2.   5. The image analysis system according to claim 1, wherein the imaging unit includes a long-focus optical system, and is a long-distance type capable of imaging the measurement object separated by a long distance. Image processing in an image analysis system that compares time-series image data obtained by imaging an object to be measured by imaging means at predetermined imaging time intervals and analyzes a velocity field of a predetermined grayscale image included in the image data A computer program comprising the means,
Filtering means comprising a high-pass filter for cutting spatial frequency components below a predetermined frequency from image data at two or more consecutive times;
PIV analysis means for converting a spatial frequency component after being filtered by the filtering means into an image, and obtaining a velocity vector between predetermined grayscale images included in the obtained image data at two or more times by PIV analysis;
Determination means for determining whether or not an error vector is included in the velocity vector of the grayscale image obtained by the PIV analysis means;
When the determination unit determines that an erroneous vector is included, the cutoff frequency of the filtering unit is adjusted, and the filtering unit executes the process of filtering the image data again, and the PIV analysis unit A computer program comprising reprocessing command means for re-executing the PIV analysis according to the above.   The determination means includes means for determining, from among the plurality of speed vectors obtained by the PIV analysis means, a speed vector showing a relatively different tendency with respect to the magnitude or direction of the speed vector as an erroneous vector. The computer program according to claim 6. The filtering means is set to perform a plurality of filtering processes at different cutoff frequencies for one image data,
The PIV analysis means obtains a velocity vector by PIV analysis between image data of the same cutoff frequency by the filtering means,
The determination means comprises means for comparing a plurality of velocity vectors obtained for each cutoff frequency and determining whether the magnitude of the velocity vector or the degree of variation in direction exceeds a predetermined threshold value. The computer program according to claim 6.

Images