JPH0635944B2 - Visualization method of flow by image processing - Google Patents

Description

TECHNICAL FIELD The present invention relates to a flow visualization method by image processing, and more specifically, enables a flow field visualization method to quantitatively determine a physical quantity of a flow field. The present invention relates to a flow visualization method by image processing.

[Conventional technology]

There have been many attempts to grasp the flow by the visualization method of the flow field, and wall trace, tuft method, direct injection method, etc. are well known.

The wall surface tracing method is a method in which oil or the like is applied to the surface of an object and the flow state, direction, speed, etc. are obtained from the streak pattern that appears due to the flow.

The tuft method is a method in which a large number of threads are stretched on the surface of an object and the flow is measured from the degree of fluttering.

The direct injection method involves placing a dye in a stream and visualizing the dye trail.

[Problems to be Solved by the Invention]

However, the wall tracing method has a problem that it is difficult to measure the flow in a space away from the surface of the object, and the tuft method makes it difficult to measure an arbitrary cross section. Further, the above direct injection method has a problem in that the velocity of the dye on the trace can be grasped, but the entire flow region cannot be visualized at one time.

In recent years, there has been remarkably strong interest in visualizing the entire flow region at one time and calculating physical quantities such as a flow velocity and a flow function, and calculating temporal changes in the physical quantity of the entire flow region.

Interest in visualizing an arbitrary cross section in a flow region and visualization of not only the velocity of water alone but also the velocity of bubbles in water, etc. is extremely strong. The Society of Mechanical Engineers Collection (B) "55
Vol. 509 (1989-1), pp. 107-114, has been proposed. In this method, tracer particles are mixed in the flow field, and the tracer particles are exposed to continuous light or strobe light to image-process the trace. In image processing, for example, an image is input from a television camera, its frame information is converted into field information, and field information for four consecutive times is subjected to image processing to trace the trajectory of each particle. Then, the flow field is visualized from the traces of the individual tracer particles to obtain various physical quantities.

Explaining the method shown in the above paper in brief, first, field information for four consecutive times is extracted from a signal input from a television camera, and image processing is performed at each time to obtain tracer particles. Find the center of gravity of. The image processing executed at this time is noise processing,
It is a binarization process and a gravity center position calculation process. That is, after removing the background image from the obtained image information, binarization is performed at a predetermined threshold value, and noise processing for removing particles smaller than tracer particles is performed, and then labeling processing is performed. The barycentric position of each particle is calculated from the constituent pixels (x _i , y _i ) of each particle obtained by the processing.

Next, the particles are associated with each other. First, the search range is designated by an appropriate radius from the position of the center of gravity of the particles to be tracked at the first time. Then, the barycentric positions of all the particles existing within this search range in the image at the second time are extracted as the barycentric positions of the destination candidates. Next, the velocity vector of the particle of each destination candidate is calculated from the position of the center of gravity of the particle at the first time and the position of the center of gravity of each destination candidate extracted at the second time. And based on those velocity vectors,
Each search center point at the third time is determined, and the search range is set around each search center point with a relatively small radius. Then, the position of the center of gravity of the particles existing in each search range is extracted as the destination candidate at the third time. Further, the destination candidates at the fourth time are extracted by the same procedure. In this way, the route in which the destination candidate exists at all of the second to fourth times is extracted as the virtual route of the particle. And
If there is one extracted virtual path, it is recognized as the movement path of the particle. On the other hand, when there are a plurality of extracted virtual routes, the most appropriate route is selected from them based on a predetermined determination condition.

The above tracking process is performed for all particles, and if there is a path that shares the same coordinates at the same time, they are deleted as defective data. In addition, the data of particles that could not be associated at all times due to reasons such as flowing out of the measurement area within the measurement time is not adopted.
A predetermined physical quantity, such as a flow velocity or a function, is obtained from the movement path of each tracer particle thus obtained.

The present invention has a problem that it is impossible to visualize any cross section in a flow region at a time, which is a drawback of conventional visualization methods such as a wall tracing method, a tuft method, a direct injection method, and the visualization information. It is intended to solve the problem that it is impossible to quantitatively determine the physical quantity of fluid such as flow velocity and stream function from the above, and the method of the above paper was improved to visualize the flow field easily. Then, it aims at proposing the flow visualization method by the image processing that made it possible to quantitatively determine the physical quantity of the flow field.

[Means for Solving the Problems]

MEANS TO SOLVE THE PROBLEM This invention for solving the said subject WHEREIN: The particle diameter of 2
Particles with a size of less than 10 mm and less than 10 mm are mixed, and laser light is two-dimensionally spread in this fluid in the form of a light sheet and continuously irradiated,
The scattered light or fluorescence of the particles is recorded on a recording medium via a camera, the flow region to be visualized is divided into a plurality of regions according to the continuity difference of the particles, and the regions are recorded on the recording medium in each region. The field information for 3 to 6 times is extracted from the obtained image information at a time interval according to the velocity of the particles in the area, and the field information is processed in each area to perform the flow once in the area. It is a method of visualizing a flow by image processing, which is characterized in that the physical quantity such as the velocity and the flow function of the fluid is measured.

Further, in a preferred embodiment of the present invention, the velocity of gas and liquid in a gas-liquid two-phase flow composed of gas and liquid or the velocity of solid and liquid in a solid-liquid two-layer flow composed of solid and liquid are determined. , Changing the binarization level in the image processing,
Also, the physical quantity of the two-phase fluid is quantitatively measured simultaneously by distinguishing the size of the mixed particles indicating the velocity of the liquid and the size of the bubbles or the solid by at least one of the image processing. To do.

[Action]

First, the flow region is continuously irradiated with a light sheet-shaped argon laser having a thickness of 3 mm or less so that an arbitrary cross section in the flow region can be visualized.

Next, in the measurement of single-phase flow of liquid or gas, particles with a particle diameter of 2 mm or less and 10 μm or more are put in a fluid, and scattered light or fluorescence of a light sheet-like argon laser irradiated on the particles is emitted from a camera. Record once on the recording medium.

Further, when the solid and liquid in the solid-liquid two-phase flow or the gas and liquid in the gas-liquid two-phase flow are simultaneously measured, the scattering intensity of the laser light is shown, and in the solid-liquid two-phase flow, the flow velocity of the liquid is shown. By pre-adjusting the mixed particles and the solid, and in the gas-liquid two-phase flow, the mixed particles and the bubbles that show the flow velocity of the liquid to be different, and changing the binarization level at the time of image processing, the solid and the liquid or the gas. And to distinguish the flow of liquid. Alternatively, by adjusting the mixed particles in advance so that the area of the mixed particles indicating the flow velocity of the liquid during image processing and the area of the solid or bubbles change, the flows of the solid and the liquid or the gas and the liquid are distinguished from each other.

Next, for example, the recorded image information is monitored, and the flow region to be visualized is divided into a plurality of regions according to the rough velocity difference of the particles. Then, the image information recorded on the recording medium is changed to 1/60 from the frame information of 1/30 seconds, for example.
The field information is divided into seconds, and the field information for three times, which has a relatively wide time interval for the slow speed area, is stored in the frame memory connected to the computer, for the fast speed area. Field information for, for example, 6 time points having a relatively narrow time interval is stored in the frame memory. These three to six pieces of information at each time are subjected to image processing at each time to calculate the center of gravity of the particle. Then, from the calculated center of gravity, the particle traces are associated with each region in the same manner as the method of the above-mentioned paper, and the velocity and the flow direction in the entire flow region are calculated.

Further, in the above method, it is also possible to obtain the temporal change in the flow region by extracting information for 3 to 6 times from relatively different times and performing image processing on them.

〔Example〕

 Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a diagram showing an example of a device configuration for carrying out the method of the present invention, and FIG. 2 is a block diagram showing a configuration example of an image processing device.

As shown in FIG. 1, an apparatus for carrying out the method of the present invention comprises an argon laser 1, an optical fiber 2, a pinhole 3, an optical lens 4, an acrylic container 5 having a thickness of 10 mm, and a C container.
CD camera 8, recording / reproducing device 9 (12), monitor 10 (11) connected to recording / reproducing device 9 (12), frame / field converter 13, computer 15, and frame memory 14 connected to computer 15 , Monitor 16
In this device, the laser light emitted from the argon laser 1 once enters the optical fiber 2, and the optical lens 4 attached to the tip of the optical fiber 2 has a thickness of 2 mm.
The following laser sheet is made and irradiated with the fluid in the acrylic container 5. The fluid used in the examples of the present invention is water.

In this case, when the laser light from the argon laser 1 is incident on the optical fiber 2, the larger the core diameter of the optical fiber 2, the less the light loss in the optical fiber. Since the thickness of the laser sheet becomes thicker, it is preferable to make the argon laser sheet thinner by providing a pinhole 3 or the like between the optical fiber 2 and the optical lens 4.

In the embodiment of the present invention, the core diameter of the optical fiber 2 is 100.
The thickness of the laser sheet is adjusted by reducing the light loss in the optical fiber to μm and providing a pinhole 3 with an inner diameter of 1 mm at a position 30 mm from the front end of the optical fiber 2 between the optical lens 4 and the optical fiber 2 at the front end. .

Example 1 (Visualization of Water Flow) The acrylic container 5 shown in FIG. 1 is filled with water, and the upper surface of the acrylic container 5 is open to the atmosphere. Water is continuously supplied to the acrylic container 5 through a pipe 6 and continuously drained through a pipe 7.

In order to visualize the flow field of water in the acrylic container 5, particles that follow the flow of water are continuously placed from the upper surface of the acrylic container 5, and the optical lens 4 and argon are applied to the cross section of the container 5 to be visualized. The laser light was continuously applied to the light sheet. Then, C in the direction orthogonal to the laser sheet surface 1A
By placing the CD camera 8 and capturing light scattered or fluorescent when the particles pass through the laser sheet surface 1A from the direction orthogonal to the sheet surface by the CCD camera 8 and continuously recording on the recording device 9. The video on the seat surface was recorded. During recording, the monitor 10 connected to the recording / playback apparatus 9 confirmed the recording state.

As the mixed particles, it is desirable to use acrylic spherical particles or particles whose surface has been surface-modified with a dye such as methylene blue to increase scattering or fluorescence intensity. In this example, the particle size without surface modification is 30 μm, and the density is 1 kg /
cm ³ particles were used.

Next, the recorded information is reproduced by the recording / reproducing device 12, and 1 /
The frame / field converter 13 divides the frame information in units of 30 seconds into the field information in units of 1/60 seconds,
The data is recorded in the frame memory 14 connected to the computer 15.
During playback, the monitor 11 connected to the recording / playback device 12
The playback status was confirmed with.

In this example, since the recording / reproducing apparatus of the NTSC system was used, of the 480 horizontal scanning lines, only 240 lines were effective because the frame information was divided into the field information.

In addition, when recording to the frame memory 14,
An image of the mixed particles was monitored through a monitor 16 connected to the computer 15, and the area was divided into a slow area and a fast area according to the rough speed of the mixed particles. Field information in units of 1/60 seconds is recorded in the frame memory 14 at intervals of 3/30 seconds for the slow area, and in 1/60 seconds for the fast area. Field information was recorded in another storage area of the frame memory 14 for 6 times at 1/30 second intervals.

Next, the information of each time recorded in the frame memory 14 is
Image processing was performed separately for the low-velocity region and the high-velocity region, and the barycentric position of the particles was calculated. The image processing used here is nozzle processing, binarization processing, and calculation of the barycentric position of particles.
The calculated center-of-gravity position was temporarily stored in the external storage device 16 of the computer 15 separately for the low speed region and the high speed region.

Next, particles were associated with the slow speed region and the fast speed region by a method similar to the method described in the above-mentioned paper, and the speed was calculated. In this example, the number of centroids of image-processed particles in the slow velocity region is about 350 at each time, and when the method of the above paper is changed to 3 time and the correspondence is performed, 270 correspondences are obtained. It was possible to attach. Further, the number of centroids of image-processed particles in a high-velocity region is about 220 at each time, and when the method of the above paper is extended to 6 times and the correspondence is performed, 120 correspondences are obtained. It was possible.

Then, the velocity of the flow in that portion was calculated from the movement path of each particle and the measurement time obtained as described above. On the other hand, the velocity at an arbitrary place in the measurement region was obtained by drawing lines at arbitrary intervals intersecting each other in the measurement region and interpolating the velocity at the intersection of the lines with a computer. As an interpolation method, for example, measurement data of three points are searched around an intersection to be interpolated, and an area function using the distance between the three points and the intersection is processed.

In this example, the measurement area is divided into 35 in the horizontal direction and 35 in the vertical direction.

FIG. 3 shows the velocity distribution at the intersection divided into the horizontal direction 35 and the vertical direction 35, and the arrows indicate the directions and the lengths indicate the distances.

FIG. 4 shows a stream function, which is another physical quantity of the stream, obtained from the velocity distribution of FIG.

Example 2 (Measurement of water + solids) Water and solids are supplied from a pipe 6 connected to an acrylic container 5 shown in FIG. 1, and drained from a pipe 7.

Particles having a particle diameter of 30 μm and a density of 1 kg / cm ³ shown in Example 1 as particles that follow the flow of water in order to measure the velocities of water and solid in a solid-liquid two-phase flow formed of water and solid, respectively. Is put from the upper surface of the acrylic container 5, while the solid particles are surface-modified and have a particle diameter of 200 μm and a density of 5 kg.
Particles / cm ³ were fed through the pipe 6.

The weight ratio of the particles representing the solid flow in the acrylic container 5 was 1/150 (solid weight / water weight).

The process up to recording in the frame memory 14 connected to the computer 15 was exactly the same as in the first embodiment.

The recording information in the frame memory 14 is the area of scattered light and the scattered intensity, and the particles showing the behavior of water and the particles showing the solid are separated. In this example, the scattering area and the scattering intensity of the particles, which show the behavior of water, were significantly different from those of the solid and could be clearly separated.

The separated information was subjected to the same treatment as that shown in Example 1 for each of the solid and the liquid, and the flow rate was obtained.

FIGS. 5 and 6 show the velocity and stream function of the solid, which were obtained simultaneously with this example and water. The water velocity and stream function were the same as in Example 1.

〔The invention's effect〕

As is clear from the above description, in the flow visualization method of the present invention, since the laser light is spread in the form of a light sheet, it is possible to measure an arbitrary cross section in the measurement region by changing the irradiation location of the laser light. is there. Further, the inside of the measurement area can be visualized at one time.

Further, the flow region is divided into a plurality of regions according to the velocity, field information is extracted at time intervals according to the velocity in each region, and image processing is performed on the field information. It is prevented that the positions of the particles are overlapped with each other so that they cannot be associated with each other, or that in a high-velocity region, an apologetic association is performed due to a large distance between the particles.

Furthermore, by appropriately adjusting the size and type of particles, or by changing the scattering or fluorescence intensity of particles in image processing, liquid and solid in a solid-liquid two-phase flow formed of liquid and solid or liquid and gas. It is possible to simultaneously measure the velocities of gas and liquid in the gas-liquid two-phase flow formed in 1.

[Brief description of drawings]

FIG. 1 is a schematic perspective view showing a configuration example of an apparatus for carrying out the flow visualization method of the present invention, FIG. 2 is a block circuit diagram showing an example of the configuration of an image processing apparatus, and FIG. 3 is a method of the present invention. Fig. 4 is a velocity distribution diagram of water obtained by Fig. 4, Fig. 4 is a flow function diagram of water in Fig. 3, and Fig. 5 is a velocity distribution diagram of solids in a solid-liquid two-phase flow obtained by the method of the present invention. FIG. 6 is a flow function diagram of the solid of FIG. In the reference numerals used in the drawings, 1 ... Argon laser 2 ... Optical fiber 4 ... Optical lens 5 ... Acrylic container 8 ... CCD camera 9, 12 ... Recording / playback device 10, 11 ... Monitor 13 ... ... frame / field converter 14 ... frame memory 15 ... computer 16 ... monitor 17 ... external storage device.

 ─────────────────────────────────────────────────── ─── Continuation of the front page (56) References JP-A-61-274238 (JP, A) JP-A-61-29729 (JP, A) JP-A-63-85326 (JP, A)

Claims (2)

[Claims] 1. A fluid is mixed with particles having a particle diameter of 2 mm or less and 10 μm or more, and laser light is two-dimensionally spread in this fluid in a light sheet shape and continuously irradiated to obtain scattered light or fluorescence of the particles. Is recorded on a recording medium via a camera, the flow region to be visualized is divided into a plurality of regions according to the speed difference of the particles, and in each region, the image information recorded on the recording medium Visualizing the flow region at a time by extracting field information for 3 to 6 times at time intervals according to the velocity of the particles, and performing image processing on these field information in each region,
A method for visualizing a flow by image processing, which comprises measuring a physical quantity such as a velocity or a flow function of the fluid. 2. A velocity of gas and liquid in a gas-liquid two-phase flow composed of gas and liquid, or a solid-liquid 2 composed of solid and liquid.
The velocities of solids and liquids in the phase flow can be calculated as 2 in the above image processing.
The two-phase fluid of the two-phase fluid is changed by changing the value level and / or by distinguishing the size of the mixed particles and the size of the bubbles or the solid indicating the velocity of the liquid from each other in the image processing. 2. The physical quantity is quantitatively measured at the same time.
A flow visualization method by the described image processing.