JP4485929B2 - Droplet observation method and observation apparatus - Google Patents

Description

  The present invention relates to an observation method and an observation apparatus for observing a flight state of a droplet discharged from a nozzle such as an inkjet head.

Observing droplets ejected from an inkjet head is used in many fields, including inspection in the head manufacturing process, performance verification in the head development process, and research on matching with ejected liquid in basic head research. It is valid. As an apparatus for observing a droplet, a device that measures the velocity (flying speed) of the droplet in conjunction with the observation of the droplet is known. The speed of the droplet is calculated by calculating the amount of movement from the droplet captured as an image by intermittently turning on and off the illumination and shooting the flying droplet intermittently. It can be calculated from Japanese Patent Application Laid-Open No. 2003-28696 discloses that a droplet ejected from an inkjet head is continuously captured by a camera, and a difference in the center of gravity of the droplet from a plurality of consecutive image data and a difference in time when the droplet is captured. Discloses a droplet amount measuring apparatus for determining a droplet velocity.
JP 2003-28696 A

  In Japanese Patent Laid-Open No. 2003-28696, the speed of a droplet is measured by utilizing the fact that a droplet discharged from a nozzle follows a predetermined path and passes a predetermined position. For this reason, the periphery of the droplet path is covered with an acrylic plate so that the path or flight path of the droplet does not change due to disturbance. Further, an electrode or a deflecting plate that forcibly applies an electric field to the droplet discharged from the nozzle is provided so that the droplet forcibly follows a fixed path.

  In an environment where droplets are actually ejected from an inkjet head, there is always a disturbance. Further, the path of the droplet is not controlled. Therefore, in order to check the performance of the head in a state closer to the actual machine and to check the matching with the discharge liquid, it is possible to prevent disturbance with a special device or method, and to control the path of the droplet after discharge. It is desirable to be able to observe the flying state of the droplet.

  However, it is difficult to automatically determine the speed of a droplet unless the position where the droplet passes through is determined. When observing a droplet or measuring a velocity, the droplet may not appear at a predetermined position if it is affected by a disturbance. Moreover, even if a droplet appears at a predetermined position, there is no guarantee that the droplet is an observation target droplet. For example, when a liquid in which fine particles are diffused is discharged from a nozzle, droplets may be discharged in various directions depending on the diffusion state of the fine particles. In particular, there is a high possibility that such a phenomenon will occur at the stage of examining and researching the matching between the head and the liquid. For this reason, in extreme cases, the liquid droplets output from multiple nozzles may intersect, or even if they do not intersect, multiple liquid droplets may pass through very close positions. There is no guarantee that the droplet is the one to be observed. Furthermore, since the shape of the droplet changes during the flight, or it is divided into a plurality of droplets, it is difficult to determine that the droplet that appears at the planned position is the droplet to be observed. There is no guarantee that you can observe a drop when you want to observe it.

  From this point of view, conventional droplet velocity measurement is a technique based on the premise that the droplets appearing at a predetermined position in the image are the same, so that the same droplet appears at the predetermined position. It can be said that the flying state of the droplet is controlled.

  Also, when shooting droplets ejected by the camera, the lens or CCD of the camera is dusty, or there is a defect or damage, and the effect of such defect on the image that shows the course or path through which the droplet passes. If there is, there is always the possibility of misrecognizing dust or defects as droplets. Although it may be possible to solve the problem of misrecognition by repeatedly performing lens cleaning and calibration to eliminate the effects of dust and defects, it takes a lot of effort to completely eliminate the possibility of misrecognition. In order to perform observation efficiently, it is desirable that false recognition can be automatically eliminated.

  Therefore, an object of the present invention is to provide a droplet observation method and an observation apparatus that can accurately capture the droplet to be observed as an image intermittently in the actual environment of the droplet path. . In the present invention, the droplet to be observed is intermittently imaged even in an environment in which disturbance is affected during the flight of the droplet, or under conditions where the flight direction of the droplet cannot be predicted. It is an object of the present invention to provide an observation method and an observation apparatus for liquid droplets that can be captured with high accuracy. In addition, there is provided a method and apparatus capable of accurately observing a desired droplet by eliminating even the cause of unpredictable misrecognition due to dust or defects in image acquisition means such as a camera for obtaining an image. This is the object of the present invention.

  For this reason, in the present invention, droplets are intermittently imaged or photographed, and a process for confirming the identity of the droplets included in the images is provided for each obtained image. The observation method of the present invention includes an image acquisition step of intermittently acquiring images of droplets ejected from a nozzle at a first time interval, and a first image and a second obtained before and after the time. A recognition step of determining the identity of the droplets included in the first and second images. In addition, the observation apparatus of the present invention includes an image acquisition unit that intermittently acquires an image of a droplet discharged from a nozzle at a first time interval, and a first image obtained before and after the time. Recognizing means for determining the identity of the droplets included in the first and second images between the second images. Further, in the present invention, a step of intermittently obtaining images of droplets ejected from the nozzles at a first time interval, and a first image and a second image obtained before and after time. And an observation program (program product) having instructions that can be executed by a computer in the recognition step of determining the identity of the droplets included in the first and second images. With this program, it is possible to configure an observation apparatus using resources of a general-purpose computer. This program can be provided by being recorded on a recording medium such as a CD-ROM or ROM, or via a computer network such as the Internet. Can be provided.

  Intermittent images taken at appropriate intervals while the droplets are flying are fixed for each image because the droplets are in the middle of flying. Easy to grasp the state. Therefore, it is possible to compare the position through which a specific portion of the droplet passes with the background target, and it is possible to accurately acquire the velocity and direction of the droplet. However, measurement of velocity and direction is meaningless unless the observed droplets are the same in the first image and the last image in which measurement of velocity and the like is started. Therefore, in the present invention, by determining the identity of the droplets included in the first and second images between the first image and the second image obtained before and after in time, It is ensured that the observed droplet is the same between the first image and the last image from which a measurement such as velocity is started.

  That is, in the observation method and observation apparatus of the present invention, it is determined that the droplets included in the first and second images obtained before and after are the same, and the droplets are the same. By making the determined second image the first image with respect to the subsequent image, it becomes possible to sequentially determine the identity of the droplets included in the subsequent images. Therefore, in the observation method and apparatus of the present invention, the droplet included in the first image to be measured without controlling the flight direction of the droplet or removing the influence of disturbance on the flight direction of the droplet. And the droplet included in the last image can be guaranteed to be the same.

  Therefore, with the observation method and observation apparatus of the present invention, the actual flight state of the droplets ejected from the nozzle can be observed, and the velocity and direction of the droplets can be measured. Furthermore, even in a situation where a plurality of droplets are ejected, it is possible to determine that the droplets included in the preceding and succeeding images are the same by paying attention to any one droplet. Therefore, it is possible to guarantee that the droplets captured at the beginning and end of the measurement are the target droplets even under conditions where multiple droplets may cross or pass through very close positions. Regardless of the flight status of the droplet, the state of the target droplet can be observed reliably, and physical properties such as velocity can be measured. In the image acquisition step or means, when the image including a plurality of droplets discharged from one or a plurality of nozzles is acquired intermittently at the first time interval, the recognition step or means uses the first Among the plurality of droplets included in the second image, it is possible to determine the identity of one or more designated droplets and automatically track the designated one or more droplets.

  Furthermore, by determining the identity of the droplets included in the images obtained before and after, the position of the droplet included in the middle of the measurement or in the last image is limited to some extent. Therefore, even if there is a noise component that may be determined as a droplet due to dust or a defect in the image, it is easy to separate the noise component and the target droplet as with other droplets. In particular, since the noise component does not move, it is possible to determine an image having no position change as noise from images that fluctuate in time, and in that respect, it is easy to separate from the target droplet, and for a short time Therefore, the possibility of misrecognition can be eliminated with high accuracy by processing with low load.

  In some cases, it is difficult to separate the noise component first appearing in the second image from the droplet. For example, immediately after starting automatic tracking of a droplet, it is difficult to limit the shape and flight direction of the droplet, and when a component that causes noise appears in the droplet to be tracked, the shape and flight direction There is a possibility that it cannot be separated by referring to the above. For this reason, in the recognition step or means, the droplet included in the first image and the droplet included in the second image by using the third image obtained after the second image in time. It is desirable to provide a process or means for judging identity. The noise component that does not move in time can be deleted from the droplet candidates of the second image by referring to the third image. Also, if the noise component is another droplet whose flight direction has changed significantly due to some factor, it is included in the range of the flight distance calculated from the position of the droplet in the first image or the search area in the third image. There is a high possibility that Therefore, such a noise component can also be deleted from the droplet candidates of the second image by referring to the third image.

  Furthermore, by determining the identity of the droplets included in the images that change in time, it is possible to reliably determine the identity of the droplets whose shape or state changes during the flight. That is, if the shape of the droplet is different at the beginning and end of the measurement, it is difficult to determine the identity even if the traveling direction of the droplet is controlled. However, even if the shape changes during flight, the liquid drops separate into two, or the state changes, the change is slight between images that change in time, and automatically changes according to appropriate criteria. It is possible to determine the identity or similarity of droplets. Therefore, according to the present invention, it can be assured that the same droplet is measured even if the shapes are greatly different at the beginning and the end of the measurement. For example, in images that change in time, if the droplets are the same, there is little change in the outer shape, area, density, and length in the vertical and horizontal directions reflected in the image. Drops can be determined to be identical and tracked.

  Once the droplet discharge method is determined, the droplet flying speed falls within a limited range. For example, in an ink jet method in which droplets are ejected from nozzles by an actuator, the velocity range of the droplets is 0.5 to 20 m / s. Therefore, from the position of the target droplet in the first image, a small area including the maximum flight distance of the interval (first time interval) for intermittent shooting or imaging is set as the second image, and the small area By confirming the presence / absence of the liquid droplets in the inside, the identity of the liquid droplets included in the first image and the liquid droplets included in the second image can be confirmed. For this reason, the recognition step or the recognition means sets an area setting for setting a search area including the range of the maximum flight distance of the droplets in the first time interval from the position of the droplets of the first image in the second image. It is desirable to provide a process or area setting means and a droplet detection process or droplet detection means for determining the presence or absence of a droplet in the search area. This determination method can reduce the processing time because it can determine the identity of the droplets to be observed by searching a small search area rather than searching the entire image, and can also reduce the processing time. It is possible to eliminate the influence of the droplet image and noise components. If a plurality of droplets in an image are automatically tracked, a plurality of search areas can be set for each droplet.

  In the area setting process or area setting means, the search area can be further limited by setting the search area in the flying direction of the droplet of the first image, and the processing time can be further reduced. It is easier to eliminate the influence of other droplet images and noise components outside the area. Furthermore, even if the droplets fly in an arbitrary direction, the same droplet can be reliably recognized as long as the target droplet is included in the next image. The flying direction of the droplet can be obtained from the position of the droplet in the two images that precede and follow. Furthermore, if the droplet flight direction is known, even if multiple droplets or noise components are included in the same search area, other droplets and noise components can be determined from the droplet flight direction. Can be removed.

  In addition, the droplet images included in the preceding and following images are not necessarily the same. Rather, as described above, the observation method and apparatus of the present invention are suitable for observing a droplet whose shape or situation changes during flight. Therefore, it is desirable that the observation method and apparatus include a similarity determination step or similarity determination means for determining whether or not the droplet candidates in the search area are similar to the droplets in the first image. It is possible to make a similar determination using information on at least one of the outer peripheral shape, area, density, vertical / horizontal length, and position (position change) of a droplet candidate, and noise can also be removed at this stage. Furthermore, even if the droplet images included in the preceding and following images are not exactly the same, it is determined that they are the same as long as they are in similar ranges. It becomes possible to trace the drops reliably.

With or without similarity judgment, other droplets that interfere with droplet observation are detected, and when such other droplets are detected, the observation is stopped and other droplets are stopped. The influence of droplets can be eliminated. Therefore, as recognized engineering is in the second image, the flying direction of the search area, set the noise search area including a range of flight direction of the end of the maximum flying distance of the droplet in a first time interval of the search area And a step of stopping the observation when there is a droplet candidate in the noise search area . Further, the recognition means of the observation device includes a noise search area including a range of the maximum flight distance of the droplet in the first time interval from the end of the search area in the flight direction of the search area in the second image. And means for stopping the observation when there is a droplet candidate in the noise search area .

  The noise search area is an area that is further outside the maximum flight distance of the droplet, does not include the droplet to be observed, and the droplet candidates in the noise search area are droplets that are not being tracked . Furthermore, since the noise search area is in front of the flight direction of the search area, the observation target droplet is likely to fly at the next timing, and at the next timing, it is recognized as a droplet candidate in the search area, or In the worst case, a droplet that is being tracked and a droplet that is not being tracked are combined at the next timing and recognized as a droplet candidate in the search area. It is sufficient if the droplets are distinguished by the similarity judgment, but an error occurs when the droplets cannot be distinguished, and an erroneous observation result is obtained if the similarity judgment is wrong. In addition, an erroneous observation result can be obtained when the droplets are integrated. Therefore, there is a very high probability that an erroneous observation result will be obtained. Therefore, by stopping the observation, the accuracy of the previous observation result can be maintained.

  It is effective to remove a component known as noise from the image together with the similarity determination or separately from the similarity determination. In the observation method of the present invention for observing droplets, if all information other than droplets on an image is considered to be basically noise, a reference image before discharging droplets from a nozzle before the image acquisition step It is effective to provide a step of acquiring the noise component, and further to remove a noise component common to the reference image from the second image before the recognition step. In addition, it is effective that the observation apparatus further includes a means for acquiring a reference image before discharging a droplet from the nozzle by the image acquisition means and removing a noise component common to the reference image from the second image. .

  In the droplet observation apparatus, it is possible to change the image acquisition interval by providing means for controlling the first time interval based on the recognition result of the recognition means.

  In the observation method and the observation apparatus of the present invention, the droplets included in the first image of the measurement and the last image are confirmed by successively confirming the identity of the droplets included in the images before and after intermittently obtained. It ensures that the contained droplets are identical. Therefore, according to the present invention, when automatically measuring a physical property value such as a velocity of a droplet discharged from a nozzle, the traveling direction of a droplet to be measured is limited, or a disturbance that should be affected in an actual machine is not affected. Without avoiding the influence, it is possible to observe the flying state of the droplet and measure various physical property values. For this reason, according to the present invention, it is possible to observe the flying state of the droplets under more realistic conditions, the results of performance inspection of the head having the nozzle, and the evaluation results such as matching with the liquid of the head in the research stage Can be obtained with higher accuracy.

  Hereinafter, the present invention will be described in more detail with reference to the drawings. FIG. 1 shows an outline of the observation system. The observation system 1 can be used in a wide variety of situations, such as a manufacturing inspection process of an inkjet head, a head performance inspection, and a basic research stage such as a head-liquid matching. The observation system 1 includes an inkjet head 2 that ejects droplets 3 and an imaging device 11 that images the droplets 3 ejected from the inkjet head 2. In this example, the imaging device 11 includes a CCD camera 15 and a light emitting unit 16 made of LEDs. The configuration of the imaging device 11 is not limited to this, and any configuration that can intermittently photograph the flying droplet 3 is acceptable.

  The observation system 1 further includes a head controller 12 that outputs a control signal (or discharge signal) φ1 that discharges the droplet 3 to the inkjet head 2, and a light emission timing controller 13 that outputs a drive signal φ2 that causes the light emitting unit 16 to emit light. Based on the image information φ3 imaged by the CCD camera 15, the flight state of the droplet 3 ejected from the inkjet head 2 is observed, the flight speed, the flight direction (or flight angle, ejection angle), the droplet 3 And a measurement unit 10 for measuring volume and the like.

  The measurement unit 10 uses a personal computer as hardware, and functions as a measurement device by installing an observation program 17 provided by a recording medium such as a CD-ROM 16. The observation program 17 can also be provided via a computer network such as the Internet.

  FIG. 2 shows a schematic configuration of the measurement unit 10. The measurement unit 10 outputs a control signal φ4 to the head controller 12 and supplies a light emission control signal φ5 to the light emission timing controller 13 and a light emission control signal φ5 to control the timing at which the inkjet head 2 discharges the liquid droplet 3 to emit light. And a light emission control unit 22 that controls the timing at which the unit 16 emits light. The light emission controller 22 can control the timing and interval (interval, first time interval) Ti at which the CCD camera 15 images the droplet 3. The light emission timing controller 13 emits the synchronization signal (drive signal) φ2 of the interval Ti set by the control signal φ5 from the timing delayed by the delay time set by the control signal φ5 with respect to the ejection signal φ1 of the head 2 16 is output. The interval Ti is normally set to several μs, and the light emitting section 16 emits light at the interval Ti, whereby an intermittent image of the droplet 3 in flight is taken by the CCD camera 15. The timing at which the droplet 3 is ejected and the timing at which the light emitting unit 16 is driven can be automatically controlled by the measurement unit 10. It is also possible to set via the input means 19 such as a keyboard of the measurement unit 10.

  The measurement unit 10 includes an image acquisition unit 23 that acquires an image of the droplet 3 ejected from the nozzles of the head 2. The image signal φ3 of the droplet 3 intermittently obtained by the CCD camera 15 is an image The image is input to the image acquisition unit 23 of the measurement unit 10 via the capture board 18. The image acquisition unit 23 acquires an image φ3 of the droplet 3 intermittently photographed at every interval Ti, and records the image before discharging the droplet from the nozzle in the memory 28 as a reference image φ9. The measurement unit 10 further includes a recognition unit 24 that determines the identity of the droplet 3 included in these images between the images φ3 obtained before and after in time, and the droplet 3 included in each image. Position information φ11, a log unit 27 that obtains information φ12 related to the form of the droplet 3 and stores it in the memory 28, and calculates the flying speed and angle of the droplet 3 based on the position information φ11 stored in the memory 28 And an analysis unit 26.

  The recognition unit 24 assumes that the intermittent image φ3 that fluctuates in time is the first image φ31 and the second image φ32, and the droplet 3 of the first image φ31 is included in the second image φ32. An area setting unit 31a for setting a search area including the range of the maximum flight distance of the droplet 3 in the interval Ti from the position; a droplet detection unit 32 for determining the presence or absence of the droplet 3 in the search area; And a similarity determination unit 33 for determining whether or not the droplet 3 is similar to the droplet 3 of the first image φ31. Further, the recognizing unit 24 sets a noise search area setting unit 31b that sets a noise search area including the range of the maximum flight distance from the end of the search area in the flight direction of the search area in the second image φ32. And a determination unit 37 for determining whether or not to stop the observation when there is a droplet candidate in the noise search area. The recognizing unit 24 includes a noise removing unit 34 that removes components included in the reference image φ9 as stationary noise from the first image φ31 and the second image φ32 with reference to the reference image φ9. Yes.

In the case of an inkjet head, the velocity range of the droplet 3 ejected from the head 2 is 0.5 to 20 m / s. Therefore, the maximum flight distance Lm is expressed by the following formula (1).
Lm = 20 × Ti (1)

  The area setting unit 31a sets a search area including the maximum flight distance Lm of the droplet 3 in the second image φ32 from the position of the droplet 3 in the first image φ31. If the interval Ti is set to 10 μs, a search area including a range of 0.2 mm from the position of the droplet 3 in the first image φ31 is set.

  Furthermore, the area setting unit 31a sets a search area in the flight direction of the droplet 3 of the first image φ31. Since the memory 28 stores the position information φ11 of the droplet 3 included in the first image φ31 and the previous image, the flight of the droplet 3 in the first image φ31 from the position information φ11. You can know the direction. Therefore, the search area is not set to a circle or a rectangle centering on the position of the droplet 3 in the first image φ31, but is set to a limited range centering on the traveling direction of the droplet 3. The

  Similarly, the noise search area setting unit 31b also sets the noise search area within a limited range including the maximum flight distance Lm beyond the search area with the traveling direction of the droplet 3 as the center.

  The droplet detection unit 32 determines the presence / absence of the droplet 3 from the area and shape of regions having different contrasts in the search area of the image φ32. The presence or absence of the droplet 3 can be automatically determined by setting the contrast between the region determined to be the droplet 3 and the background of the observation screen as a threshold value in the droplet detection unit 32 in advance.

  The similarity determination unit 33 compares the information φ12 regarding the form (shape, area, etc.) of the droplet 3 of the first image φ31 obtained previously with the form information φ12 of the droplet 3 of the second image φ32, and It is determined whether or not the droplet of the image φ31 is the same as the droplet determined as the droplet in the search area of the image φ32. In the similarity determination unit 33, if the shape information φ12 of the droplet 3 in the first image φ31 is similar to the shape information φ12 of the droplet 3 in the second image φ32, the droplets 3 are the same. It is judged that. Therefore, even if the shape of the droplet 3 included in the first image φ31 and the shape of the droplet 3 included in the second image φ32 are not exactly the same, the same droplet 3 can be tracked or tracked, It can also track droplets that may change shape or separate during flight.

  The log unit 27 includes a position information generation function 38 that obtains a selected location of the droplet 3 in the search area as the position information φ11, for example, the coordinates of the tip or the center of gravity of the droplet, the outer peripheral shape of the droplet 3, and the droplet 3 vertical (Y direction) and horizontal (X direction) lengths, and their average values, the area of the droplet 3 and the concentration of the droplet 3, the position of the droplet 3, or a plurality of morphological information and a form information generation function 39 for obtaining φ12. Although the determination of the droplets can be performed using at least one of the morphological information φ12, combining a plurality of pieces of information can increase the accuracy of similarity determination, although the processing burden increases.

  The analysis unit 26 first includes the droplet 3 at a position closest to the first position (first droplet arrival point) and the second position (second droplet arrival point). The velocity and angle of the droplet 3 are calculated from the first image and the last image. The first droplet arrival point and the second droplet arrival point can be set from the input means 19 such as a keyboard. For example, the first droplet arrival point is set to 0 to 0 from the nozzle surface of the head 2. About 2 mm, the second droplet arrival point is set in a range of about 0 to 0.5 mm from the surface where the droplet reaches (a range of about 0.5 to 2.0 mm from the nozzle surface of the head 2). It is possible. It is desirable that the image for calculating the velocity includes the nozzle surface that is the base of the droplet arrival point or the surface that the droplet reaches together with the droplet. By measuring the distance from the reference surface, the accuracy can be improved. High analysis is possible.

  The measurement unit 10 further includes a display control unit 29 that displays on the LCD 20 the image φ3 obtained by the image acquisition unit 23, the analysis status, and the like.

  FIG. 3 is a flowchart showing a process of measuring the droplet velocity in the measurement unit 10. FIG. 4 is a flowchart showing the operation of automatically tracking a droplet in the process of measuring the velocity. FIG. 5 shows an image obtained for each interval Ti immediately after the droplet 3 is ejected from the nozzle 2a, and the processing of FIGS. 3 and 4 will be described with reference to these images. These droplets 3 transferred from FIG. 5A to FIG. 5G are columnar immediately after being ejected from the nozzle 2a of the head 2, but thereafter, due to the influence of the surface tension of the liquid itself. The shape changes, and finally it flies in the Y direction while being separated into the main droplet 3a and at least one satellite 3b.

  In the process of measuring the velocity of the droplet 3, in step 51, designation of the first droplet arrival point H1, the second droplet arrival point H2, the imaging interval (interval Ti), and the form for determining the droplet 3 is performed. Observation conditions including the above are set. When the observation of the droplet 3 is started in step 52, first, in step 90, the reference image φ 9 before the droplet is ejected from the nozzle is captured and recorded in the memory 28. Next, in step 53, an image of the droplet 3 ejected intermittently from the nozzle at the interval Ti is acquired, and the droplet 3 is automatically tracked. If it is determined in step 54 that the position of the automatically tracked droplet 3 has reached the first droplet arrival point H1, as shown in FIG. Time is recorded. Furthermore, if it is determined in step 56 that the droplet 3 has automatically reached the second droplet arrival point H2 as shown in FIG. The tracking is completed, and the flying speed and ejection angle (or straightness) of the droplet 3 are obtained from the time and position at that time.

  As shown in FIG. 4, in the automatic tracking, first, when the interval Ti elapses in step 61 and the light emitting unit 16 of the LED emits light, in step 62 the light emitting unit 16 emits light and an image of the droplet 3 is displayed. Take a picture. In the measurement unit 10, the image acquisition unit 23 acquires the image φ3, and the noise removal unit 34 of the recognition unit 24 compares the acquired image data for inspection or observation with the reference image φ9, and refers to the reference from the inspection image. The static noise component recorded in the image φ9 is removed. By this processing, it is possible to remove noise components that do not move due to the elapsed time such as the background included in the inspection image, the inherent pixel error of the CCD camera 15, or scratches and dirt on the lens surface. Thereby, the recognition unit 24 can extract only the dynamic image component and proceed with the subsequent processing, and can basically capture only the droplet. This static noise removal can be performed for the entire image or for each search area shown below, and it is easy to leave static information necessary for observation of the arrival points H1 and H2, etc. is there.

  Next, in step 63, the area setting unit 31a of the recognition unit 24 sets an area where the droplet 3 is searched. If the measurement unit 10 acquires the image 41b in FIG. 5B as the second image φ32, the area setting unit 31a sets the previous image 41a in FIG. 5A as the first image φ31, A search area 35 having a maximum flight distance Lm in the droplet flying direction from the position of the droplet 3 included in the image 41a is set in the image 41b. In this example, the droplet 3 is traveling straight on the screen 41a in the vertical (vertical, Y) direction. Therefore, as shown in FIG. 6A, the area setting unit 31a extends from the position of the droplet 3 in the image 41a by the maximum flight distance Lm in the Y direction, and the droplet (3 in the horizontal (horizontal, X) direction). A rectangular search area 35 having a symmetrical width as a center is set. The search area 35 may be a semicircular area having a radius Lm centered on the position of the droplet 3 of the image 41a. However, the flying direction of the droplet 3 does not change rapidly, and the processing time for detecting the droplet 3 is shorter when the area of the search area 35 is smaller. Therefore, in this example, a rectangular search area 35 having a certain width (for example, the maximum flight distance Lm) in the flight direction of the droplet 3 is set. If the flight direction of the droplet 3 in the image 41a is inclined in the ± X direction in step 63, the area setting unit 31a determines the ± X direction from the position of the droplet 3 in the image 41a according to the inclination. A search area 35 biased to is set. Thereby, the search area 35 can be set so as to reliably cover the flight direction of the droplet 3.

  In step 81, the noise search area setting unit 31b sets the noise search area 36 ahead of the flight direction of the search area 35, as shown in FIG. This noise search area 36 is intended to determine the presence or absence of noise that obstructs the observation of the droplets that have entered the tracking area 35 and tracked at the next timing. It is desirable that the length in the flight direction is about 1 to 10 times the maximum flight distance Lm and the width direction is about 1 to 10 times the maximum flight distance Lm with respect to the droplet search area 35. Increasing the noise search area 36 increases the possibility of detecting droplets other than the tracking target droplets or other noise, and further increases the load of image processing. On the other hand, if the area of the noise search area 36 is too small, there is a possibility that noise entering the liquid droplet search area 35 cannot be detected in advance. Therefore, it is more desirable that the noise search area 36 has the same length in the flight direction as the flight search area 35 or about 1 to 3 times the length, and the length in the width direction orthogonal to the flight direction is also the same. The noise search area 36 does not necessarily need to be set to be continuous with the droplet search area 35, but detects a droplet with a large difference in velocity in advance or detects the droplet search area 35. It is desirable that the search areas 35 and 36 are continuous in order to accurately detect noise that enters at the next timing. The search areas 35 and 36 are not limited to a square such as a square or a rectangle, but may be a sector. When the area of the noise search area 36 is set larger than the search area 35, the resolution of image analysis can be reduced, but the noise detection rate is lowered.

  Next, in step 82, the determination unit 37 analyzes the image of the noise search area 36 and determines that there is a droplet candidate in the noise search area 36. In step 65, the droplet tracking is impossible. Judgment and stop observation. If static noise has been removed in step 80, the noise determined in step 82 is basically dynamic noise. The dynamic noise is most likely other droplets that have crossed or approached the tracking droplet for some reason. Therefore, as shown in FIG. 7A, there is another droplet 4 in the flight direction of the tracked droplet 3, and as shown in FIG. 7B, another droplet is present in the noise search area 36. When 4 is detected, as shown in FIG. 7C, there is a high possibility that another droplet 4 will enter the droplet search area 35 at the next timing. In that case, even if a similar determination as shown below is made, there is a possibility that it cannot be clearly differentiated, and there is a possibility of erroneous recognition. Furthermore, there is a possibility that the droplet recognized in the noise search area 36 will merge with the droplet being tracked at the next timing. Therefore, in step 82, if there is a possibility of erroneous recognition, the observation is stopped (interrupted) in advance to ensure the reliability of the data so far.

  The process of setting the noise search area 36 shown in steps 81 and 82 and stopping the observation when a droplet candidate is found in the noise search area 36 is one filtering function, and is based on the similarity determination described below. It may be applied simultaneously with the filtering function, or only one of them may be applied. Also, instead of suspending the observation, record the timing when the noise was found in the noise search area so that the previous observation record and the subsequent observation record can be analyzed separately. good.

  If no droplets that cause noise are detected in step 82, whether or not there is a video that can be determined as droplet 3 by image processing of the set search area 35 by the droplet detection unit 32 of the recognition unit 24 in step 64. Judging. If the droplet 3 is not found in the search area 35 in step 64, a detection error occurs in step 65, and automatic tracking ends. When an image that can be determined as the droplet 3 is found in step 64, the similarity determination unit 33 of the recognition unit 24 extracts the shape information φ12 of the image determined as the droplet 3 in step 66, and the previous image 41a is extracted. The shape information of the droplet 3 is compared with φ12. If the difference between the shape information φ12 of the droplet 3 in the new image 41b and the shape information φ12 of the droplet 3 in the previous image is within an allowable range, the similarity determination unit 33 detects the liquid found in the new image 41b. It is determined that the droplet 3 is the same as the droplet 3 in the previous image 41a. If it is not determined in step 66 that the droplets 3 are the same, a detection error occurs in step 65, and automatic tracking ends.

  In step 67, it is further determined whether or not the number of droplets 3 determined to be similar or identical is determined to be one. In step 67, if there are two or more candidate droplet images, in step 68, an effort is made to determine the droplet using the third image φ33 acquired at a timing subsequent to the second image φ32. I do. For example, immediately after starting the automatic tracking of a droplet, the direction of the droplet may not be limited in detail only by the first image (first image) φ31 and the next image (second image) φ32. There is sex. Then, there is a noise component similar to the droplet in the second image φ 32, and the noise component may enter the search area 35. In this case, it is difficult to determine that the noise component is not a candidate only by the expected flight direction of the droplet 3. However, if the third image φ33 next to the second image is used, the noise component does not move. Therefore, the noise component can be removed from the droplet candidates of the second image φ32, and automatic processing is performed in step 67. One droplet to be tracked can be determined.

  Since the image acquisition unit 23 of the measurement unit 10 acquires the image φ3 of the droplet 3 intermittently photographed at every interval Ti as needed, acquiring the third image φ33 following the second image φ32 is not possible. Since it is easy and there is almost no time difference therefor, using the third image φ33 does not hinder the observation of the droplet. Therefore, the noise removing unit 34 of the recognizing unit 24 can remove noise in more detail using the result of the similarity determining unit 33 and the third image φ33. That is, the noise removal unit 34 sets a search area 35 in the third image φ33 that covers the flight distance from the position of the droplet of the first image φ31 to the time when the third image φ33 was acquired. By observing the noise component, the non-moving component included in the second image φ32 and the third image φ33 can be removed as a static noise component such as dust or a scratch on the CCD.

  Since this process is common to the noise removal in step 80 described above, it is highly possible that static noise can be removed by performing either one of them. When the third image φ33 is used, processing capability is required, but there is a merit in that static noise generated due to some factor can be removed after droplets are ejected. Further, not only static noise components but also droplets output from other nozzles can be easily removed by using the third image φ33. If the droplet included in the search area of the second image φ32 is noise, the flying direction and speed are different from those of the droplets subject to automatic tracking, and therefore included in the search area of the third image φ33. This is because there is a high possibility of not. When the droplets included in the second image φ32 are determined, the second image φ32 is used as the first image, and the third image φ33 is used as the second image, and automatic tracking is performed by the method described above. Can be further performed.

  In step 67, when it is determined that the droplet 3 detected in the image 41b is the same as the droplet 3 recognized as the tracking droplet in the previous image 41a, in step 69, the log unit 27 Information on the droplet 3 detected in 41 b is stored in the memory 28. First, from the image of the droplet 3 detected by the morphological information generation function 39, the outer shape of the droplet, the vertical and horizontal lengths, the area, the density, and the like are extracted and stored in the memory 28 as the morphological information φ12. If the form information φ12 is the same as the form information used by the similarity determination unit 33 in step 66, the form information generation function 39 can be shared with the function of the similarity determination unit 33. Next, the position information generation function 38 obtains the coordinates of the portion of the droplet 3 set in advance based on the form information φ12 and stores it in the memory 28 as the position information φ11 of the droplet 3. If the position of the tip of the droplet 3 is used as the position information of the droplet, the position of the tip can be obtained by analyzing the outer peripheral shape of the shape information φ12.

  When the log of the target droplet 3 included in the image 41 b is recorded in the memory 28, the measurement unit 10 uses the function of the analysis unit 26 in step 70 to use the data of the droplet of the previous image 41 a and the subsequent droplet. Compare the droplet data in image 41b and check some relationships. First, the traveling direction of the droplet 3 is determined by comparing the position information φ11 of the target droplet 3 included in the preceding and succeeding images. Therefore, when the next image 41c shown in FIG. 5C is acquired, the data φ13 that determines the direction in which the search area 35 is set can be acquired and stored in the memory 28. Further, by comparing the shape information φ12, the degree of change in the shape of the droplet 3 included in the preceding and succeeding images can be known. If the change in the shape of the droplet 3 is significant, the identity of the droplet may not be confirmed in the next image 41c. Therefore, the analysis unit 26 determines that the interval Ti needs to be shortened in step 71, and in step 72, the light emission control unit 22 changes the setting of the timing controller 13 via the control signal φ5. When the position information φ11 of the droplets in the preceding and succeeding images is compared, if the distance is close to the maximum flight distance Lm due to some factor, the interval Ti is similarly shortened. Image 41d (see FIG. 5 (d)), image 41e (see FIG. 5 (e)), image 41f (see FIG. 5 (f)), image 41g (see FIG. g)).

  In the measurement unit 10, by automatically tracking the droplet 3 for each image, the droplet that has reached the first droplet arrival point H <b> 1 at the time of measuring the velocity of the droplet or in the flight direction. It can be guaranteed that the target droplet 3 in the included image 41a and the target droplet 3 in the image 41g including the droplet 3 reaching the second droplet arrival point H2 are the same. Furthermore, even when the flight direction of the droplet 3 cannot be predicted, the flight direction of the droplet 3 is traced, so that the image of the droplet 3 can be reliably captured and the velocity and the flight direction of the droplet can be measured. . For this reason, it is possible to observe the droplets discharged under the same conditions as the actual machine without covering the path of the droplets or applying an electric field to prevent disturbance, confirming the actual behavior of the droplets, And the flight direction can be measured.

  In addition, by setting an appropriate search area 35 for each image and recognizing the droplet, the probability of automatically recognizing the droplet can be improved and the processing time can be shortened. Can be automatically observed. And by auto-tracking, the area for recognizing droplets in the image is limited, so even if there is noise such as CCD defects in the image when automatically recognizing droplets, it can be affected Sex is very small. In addition, even when a plurality of droplets are included in the image, the image of the droplets taken outside the search area is not subject to automatic tracking and is therefore excluded from the observation target. Therefore, according to the present invention, it is possible to perform automatic measurement with high accuracy in the speed of the droplet and the flight direction.

  Further, when a plurality of droplet candidates are found in the search area 35 in step 64, the similarity determination unit 33 determines the identity of the droplet candidates with the target droplet in step 66. You may make it do. For example, if the deviation (difference or speed) of the position of the target droplet 3 for each image is known, among the droplet candidates in the search area 35, the same deviation as the droplet of the previous image is obtained. Candidates other than those recognized in the vicinity of the position can be excluded. The droplets move, and noise due to dust and camera abnormalities do not move. Therefore, the droplets and noise can be easily separated from the position information of the droplet candidates. Further, if the flight direction of the target droplet 3 for each image is known, among the droplet candidates in the search area 35, candidates recognized in the vicinity of the same flight direction as the droplet of the previous image. Other than can be excluded. Furthermore, by using the shape information φ12, it is possible to exclude other than droplet candidates similar to the droplet shape information φ12 of the previous image.

  Thus, even if noise such as dust is found inside the search area 35, it can be removed almost completely by determining the identity of the droplets inside the search area 35 as well. Furthermore, even if a plurality of droplets are found in the search area 35, the target droplets can be traced reliably. For example, if droplets ejected from a plurality of nozzles are included in the same image, the traveling direction of the droplets may cross depending on the ejection direction. By judging the identity of the droplets, the crossed droplets can be completely separated, and the target droplets can be traced reliably. In addition, in the case of a droplet in which the satellite follows the main droplet, it is possible to track only the main droplet or only the satellite as a target and measure the speed and flight direction of each.

  In addition, when the head includes a plurality of nozzles, a plurality of droplets are ejected back and forth, and an intermittent image including the plurality of droplets is obtained. In addition, even when focusing on one nozzle, when performing density gradation, a plurality of liquid droplets are ejected within a short time by a single ejection instruction. Images can be obtained. In the observation method and observation apparatus according to the present invention, all or a plurality of droplets included in an image obtained intermittently by the method and means disclosed above, or one or more of those designated droplets. Can be automatically tracked as described above. For example, it is possible to perform automatic tracking individually by setting a search area for each droplet designated by the area setting unit 31a and making a similarity determination for each droplet.

  In the area setting unit 31a exemplified above, the search area 35 extending from the position of the droplet 3 in the image 41a by the maximum flight distance Lm in the Y direction is set. However, considering that the liquid droplets always move, a minimum flight distance Ls (0.5 × Ti in the above example) is set, and as shown in FIG. It is also possible to set the position from the position 3 to the range from the minimum flight distance Ls to the maximum flight distance Lm. Since the search area 35 can be narrowed by considering the minimum flight distance Ls, there is a possibility that the processing speed can be increased. However, when there is a case where the droplet does not reach the minimum flight distance Ls due to disturbance or the like, the observation method includes a possibility that the droplet cannot be automatically tracked. The setting of the minimum flight distance Ls and the maximum flight distance Lm can be opened to the user, and the user can set it in consideration of conditions such as the type of droplet to be observed or the processing speed of the measuring apparatus. Is also possible.

  In the above example, the processing in the measurement unit 10 is provided as an observation program 17 having instructions for executing the steps of the flowcharts shown in FIGS. 3 and 4 and executed using hardware resources of a personal computer. This program product can be provided by being recorded on a suitable recording medium such as a CD-ROM or via a computer network such as the Internet. The measurement unit can also be provided as a data processing device with hardware logic to execute each step, and the processing speed can be further improved, so that measurement suitable for on-line or on-time observation and measurement is possible. Unit can be provided.

It is a figure which shows schematic structure of the observation system to which this invention is applied. It is a figure which shows schematic structure of a measurement unit. It is a flowchart which shows the process in which a measurement unit measures speed. It is a flowchart which shows the process which tracks a droplet automatically in a measurement unit. It is a figure which shows a mode that the droplet discharged from the nozzle is flying. It is a figure which shows the state which sets a search area ahead of a droplet. It is a figure which shows a mode that a noise search area is set to the flight direction of a search area, and the droplet which is not tracked is detected in a noise search area.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Observation system 2 Head 2a Nozzle 3 Droplet 10 Measurement unit 23 Image acquisition part 24 Recognition part 26 Analysis part 31a Area setting part 31b Noise search area setting part 32 Droplet detection part 33 Similarity determination part 34 Noise removal part 35 Search area 36 Noise search area

Claims (19)

An image acquisition step of intermittently acquiring an image of a droplet discharged from the nozzle at a first time interval;
Between the temporally first of the image and the second of the images obtained before and after, it has a recognition step of determining the identity of the droplets contained therein first and second images,
In the recognition step, the maximum flight distance of the droplets in the first time interval from the position of the droplets in the first image in the flight direction of the droplets in the first image in the second image. An area setting process for setting a search area including a range;
A droplet detection step of determining the presence or absence of a droplet in the search area;
A noise search area including a range of the maximum flight distance of the droplets in the first time interval from the end of the search area in the flight direction is set in the flight direction of the search area in the second image. Process,
And a step of stopping the observation when there is a droplet candidate in the noise search area . The observation method according to claim 1 , wherein the recognition step further includes a similarity determination step of determining whether or not a droplet candidate in the search area is similar to a droplet of the first image. 3. The observation method according to claim 2 , wherein in the similarity determination step, the similarity determination is performed using information on at least one of an outer peripheral shape, an area, a density, a vertical and horizontal length, and a position of the droplet candidate. In any one of Claim 1 thru | or 3, The process of acquiring the reference image before discharging a droplet from the said nozzle before the said image acquisition process,
An observation method further comprising a step of removing a noise component common to the reference image from the second image before the recognition step. In any one of claims 1 to 4, observation method further comprising the step of calculating a flight speed and / or flight angle of the droplet from the image of the droplet intermittently acquired before giving up observation. In any one of Claims 1 thru | or 5, in the said image acquisition process, the image containing the several droplet discharged from the one or several said nozzle is intermittently acquired at the said 1st time interval,
In the recognition step, an observation method of determining the identity of one or a plurality of designated droplets among a plurality of droplets included in the first and second images. 7. The recognition step according to claim 1, wherein the recognition step includes a step of determining the identity of the droplet using the third image obtained after the second image in terms of time. Observation method. Image acquisition means for intermittently acquiring images of droplets ejected from the nozzle at first time intervals;
Between the temporally first of the image and the second of the images obtained before and after, it has a recognition means for determining the identity of the droplets contained therein first and second images,
The recognizing unit is configured to determine a maximum flight distance of the droplet in the first time interval from the position of the droplet of the first image in the flight direction of the droplet of the first image in the second image. An area setting means for setting a search area including a range;
Droplet detection means for determining the presence or absence of droplets in the search area;
A noise search area including a range of the maximum flight distance of the droplets in the first time interval from the end of the search area in the flight direction is set in the flight direction of the search area in the second image. Means,
An observation apparatus comprising: means for stopping observation when there is a droplet candidate in the noise search area . 9. The observation apparatus according to claim 8 , wherein the recognition unit further includes a similarity determination unit that determines whether or not a droplet candidate in the search area is similar to the droplet of the first image. The observation apparatus according to claim 9 , wherein the similarity determination unit determines similarity using information on at least one of an outer peripheral shape, an area, a density, a vertical and horizontal length, and a position of the droplet candidate. In any one of Claims 8 thru | or 10 , The said image acquisition means further acquires the reference image before discharging a droplet from the said nozzle,
An observation apparatus further comprising means for removing a noise component common to the reference image from the second image. In any of claims 8 to 11, further comprising observing apparatus means for calculating a flight speed and / or flight angle of the droplet from the image of intermittently obtained the droplets before giving up observation. The observation apparatus according to claim 8 , further comprising a unit that controls the first time interval based on a recognition result of the recognition unit. In any of claims 8 to 13, wherein the image acquiring means, intermittently acquires an image including a plurality of liquid droplets discharged from one or more of the nozzles at the first time interval,
The recognizing unit is an observation apparatus that determines the identity of one or a plurality of designated droplets among a plurality of droplets included in the first and second images. 15. The recognition means according to claim 8 , further comprising means for judging the identity of the droplet by using the third image obtained after the second image in terms of time. Observation equipment. An image acquisition step of intermittently acquiring an image of a droplet discharged from the nozzle at a first time interval;
A computer can execute a recognition step for determining the identity of droplets included in the first and second images between the first and second images obtained before and after in time. have a Do command,
In the recognition step, the maximum flight distance of the droplet in the first time interval from the position of the droplet of the first image in the flight direction of the droplet of the first image in the second image. An area setting process for setting a search area including a range;
A droplet detection step for determining the presence or absence of a droplet in the search area;
A noise search area including a range of the maximum flight distance of the droplets in the first time interval from the end of the search area in the flight direction is set in the flight direction of the search area in the second image. Process,
An observation program including a step of suspending observation when there is a droplet candidate in the noise search area . 17. The instruction according to claim 16 , wherein in the recognizing step, a computer-executable instruction for executing a similarity determination step for determining whether or not a droplet candidate in the search area is similar to a droplet in the first image is further included. A prepared observation program. In Claim 16 or 17 , the step of acquiring a reference image before discharging droplets from the nozzle before the image acquisition step;
An observation program further comprising a computer executable instruction for removing a noise component common to the reference image from the second image before the recognition step. 19. The instruction according to any one of claims 16 to 18, wherein the step of calculating a flying speed and / or a flying angle of the droplet from the image of the droplet intermittently acquired until the observation is stopped can be executed by a computer. Furthermore, the observation program that you have.

Images