JP3676878B2 - Pollen automatic collection and analysis system - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a pollen automatic collection / analysis system that automatically collects pollen floating in the air and automatically determines its type and number.
[0002]
[Prior art and problems to be solved by the invention]
Conventionally, in early spring, the number of people suffering from hay fever is increasing, and the number is said to be 7-9% of the Japanese population.
[0003]
Various preventions and treatments have been attempted for this hay fever, but accurate dissemination information and surveys of airborne pollen are required from the viewpoint of social preventive medicine.
[0004]
Since 1987, in the Kanto region, the Metropolitan Sanitation Bureau and the Meteorological Association have collaborated to forecast weather conditions such as weather, temperature, wind direction, and wind speed based on pollen scattering measurements from survey stations. Improvements are made year by year.
[0005]
However, at present, all of this data collection depends on human labor, which is very laborious and time consuming, and results in considerable differences depending on the skill of the speculum.
[0006]
For this reason, forecasts have been made mainly on Japanese cedar pollen, which has many victims, but there are more than 10 types of airborne pollen that cause hay fever. Was difficult.
[0007]
In addition, the data on which pollen information is created in this way is created manually, and the current situation is that the collection method and measurement method differ depending on the observer, and the data is not uniform or reliable.
[0008]
The present invention automates the collection of pollen, speculum, and analysis in order to solve the problems associated with the collection of airborne pollen information. The purpose is to develop a system that can obtain this information in a short time.
[0009]
[Means for Solving the Problems]
The gist of the present invention will be described with reference to the accompanying drawings.
[0010]
A collecting device for automatically collecting pollen floating in the air, staining embedding sealed by staining pollen collected by the collecting device and the distinction and the pollen and dust so that the image easy analysis It consists of an embedding device and an image analysis device that measures the type and number of the pollen by analyzing the image based on the microscopic information of the pollen dyed and embedded by the dyeing and embedding device , Air is sucked into the device, passed through a series of collection chambers connected in series, and pollen in the air is collected in a cover glass placed in each chamber. The speed is configured to be variable, and the staining apparatus automatically transfers and embeds pollen collected by the collection apparatus on a slide glass, and automatically performs staining for identifying pollen. Further, the image analysis device is configured as follows: A magazine with a plurality of slide glasses with pollen stained and embedded by the staining device is installed in a microscope equipped with a CCD camera, and the slide glass is automatically inserted and ejected sequentially into the microscope XY stage. The XY stage is configured to sequentially move according to programming, transfer an image to an image analysis computer, identify the type by a programming criterion, and measure the number of pollen automatic. It relates to a collection / analysis system .
[0011]
DETAILED DESCRIPTION OF THE INVENTION
An embodiment of the present invention (how to carry out the invention) considered to be suitable will be briefly described with reference to the drawings, showing its effects.
[0012]
As shown in FIG. 17, the collection device for automatically collecting pollen floating in the air and the pollen collected by the collection device are distinguished from dust and dust so that the pollen is easily dyed for image analysis. If the system is composed of a dyeing and embedding device and an image analysis device that measures the type and number of pollen dyed and embedded by this dyeing and embedding device, the pollen floating in the air is automatically collected by the collecting device. The pollen can be dyed and embedded so that it can be easily analyzed with a dyeing and embedding device, and the type and number of pollen can be determined with an image analysis device. It can be measured and a series of operations from pollen collection to analysis can be automatically performed.
[0013]
Therefore, when pollen is collected manually and pollen is analyzed manually as in the past, there is a difference in the pollen collection method and analysis method, leading to a situation where the data is not uniform and reliable. It is possible to solve this problem, and it is possible to perform analysis with high reliability and consistency in a short time without the need for specialized skills for collecting and analyzing pollen, and without having to manually handle pollen types and scattering information. it can.
[0014]
As shown in FIGS. 2 and 17, the collection device sucks air into the device, passes through a plurality of collection chambers connected in series, and pollens in the air are placed on the cover glasses arranged in each chamber. When configured to collect, pollen with different particle sizes can be collected with separate cover glasses, and small garbage such as dust can be collected with cover glasses arranged in the rear collection chamber, The cover glass placed in the front collection chamber where pollen is collected is free from dust and image analysis can be performed accurately.
[0015]
As shown in FIG. 5, if the staining device is a device that transfers and embeds pollen collected by the collection device on a slide glass and automatically performs staining for identifying pollen, the collected pollen is By transferring and embedding onto a slide glass and automatically dyeing it, the dust and pollen contained in the pollen collected by the collection device are accurately identified in a short time without human intervention. be able to.
[0016]
As shown in FIGS . 6, 7, 8, and 17, the image analysis apparatus uses a magazine in which a plurality of slide glasses in which pollen is dyed and embedded by a staining apparatus is inserted in a microscope equipped with a CCD camera. The slide glass is automatically inserted into and out of the microscope XY stage, ejected, moved to the XY stage according to programming, and moved to the image analysis computer. If it is configured to identify and measure the number, it is possible to eliminate useless movement of the XY stage, and to automatically recognize the type and number of pollen in a short time and automatically count. Therefore, it is possible to perform image analysis with reliability and uniformity in a short time without requiring specialized skills to analyze pollen and manipulating pollen types and scattering information.
[0017]
【Example】
Specific embodiments of the present invention will be described with reference to the drawings.
[0018]
In this example, a collector (automatic collection device) that automatically collects pollen at a set time, and a stainer (automatic dyeing / sealing) that performs processing for spectroscopic examination of only pollen using an automatic machine. Embedding device), microscope data auto-acquisition device, and image analysis device for obtaining pollen type and particle size distribution at high speed from the acquired data.
[0019]
(a) Automatic pollen collector The pollen collector was configured as shown in FIG. 1 based on a cascade impactor collector that has been used as a manual collector.
[0020]
The air is absorbed by a pump, and pollen is collected in a cover glass with an adhesive thin film in the collector body.
[0021]
As shown in Fig. 2, the collector body is connected to four stages with different air inlet nozzle diameters (9.0 to 1.5 mm). It can be collected.
[0022]
As shown in FIG. 3, the structure of each stage is equipped with a feeding cassette containing 10 cover glasses at the bottom of the air inflow part and a storage cassette for storing the cover glass collecting pollen, and the movement of the cover glass is performed via a gear. This was done with two stepping motors.
[0023]
For control, a 96-bit digital input / output board PCN-7098 was used.
[0024]
A flowchart of the basic operation of control is shown in FIG.
[0025]
The overall schematic configuration of the present embodiment is shown in FIG. 17, and will be described below based on this.
[0026]
In the "Cascade Impactor" machine that sucks pollen, there is a thin glass called "cover glass" for attaching pollen to observe. Since the cover glass is automatically changed after a certain period of time, observation can be continued for a long time.
[0027]
The nozzles in the cascade impactor can be changed in thickness for each stage.
[0028]
By changing the speed at which the air flows, it is a device that allows pollen of any size to adhere to the cover glass.
[0029]
The pollen captured by the “automatic pollen collector” is colored by the “automatic dyeing / embedding device” so that it can be easily distinguished from dust and dust, and then the number and type of the pollen are observed by the “image analyzer”. .
[0030]
Each operation is automatically performed by computer control.
[0031]
Currently, there are various methods for collecting pollen. This method sucks the atmosphere with a pump and collects it in a cover glass with glycerin.
[0032]
In this method, different nozzle diameters are connected to four stages, and pollen with different particle diameters can be collected separately.
[0033]
Therefore, small dust such as dust can be collected at the rear stage, and there is an advantage that the front stage where pollen is collected does not get dust and image analysis is easy. The body was made of aluminum alloy to reduce weight. The drive board was prototyped using P TD62803.
[0034]
This board can operate up to six stepping motors with one board.
[0035]
Since this prototype has eight stepping motors, two in each stage, two were used.
[0036]
The I / O board used was a parallel IO board with four 8255s on it.
[0037]
Control was performed using NEC9801VM.
[0038]
In addition, as a result of comparison using a conventional manually scanned cascade impactor and Han's pollen, the number of pollen collected by the prototype exceeded that of the cascade at a suction flow rate of 7 Little / min or more. The collection capacity was low when the flow rate was low. This was due to air leakage from the bearing. This tendency was more prominent in the subsequent stage than in the previous stage.
[0039]
However, if the flow rate is increased too much, the collision speed with the cover glass becomes faster and the pollen shape is broken, which is different from the floating state.
[0040]
Therefore, it is necessary to increase the collection capacity at a low flow rate.
[0041]
In order to improve these disadvantages, as shown in FIG. 3, it is improved to a closed quadruple pollen collector using two step motors at each stage, and the bearing sliding portion is reduced, and the sequence control is performed. For this purpose, a control program for these step motors was developed and operated. As a result, as shown in FIG. 10, even in the second and third stages, the collection ability was not inferior to that of the cascade impactor.
[0042]
It was also confirmed that the operation of each part had sufficient reliability according to the program specifications.
[0043]
(b) Automatic dyeing and embedding device [0044]
What is collected on the cover glass by the collector described above contains a lot of dust in addition to pollen.
[0045]
In order to distinguish and pollen these from pollen, it is necessary to dye the pollen using a specific staining solution. At the same time, it is necessary to fix pollen.
[0046]
Therefore, the configuration of this device is to add one drop of the GV glycerin jelly staining solution in which the pollen on the cover glass stored in the above-mentioned storage cassette is kept at 90 ° C. on the slide glass, so that air does not mix into the dropped one. It is equipped with a mechanism for moving and fixing the pollen that has been superposed and dyed onto the slide glass, and a function for feeding and storing these glasses to the cassette.
[0047]
FIG. 5 shows a flowchart of the control operation of the dyeing device of this embodiment.
[0048]
Since image analysis of pollen is performed based on microscopic information of collected pollen, an observation preparation must be prepared.
[0049]
At that time, in order to easily identify the pollen, a liquid to be dyed is put in the embedding material, and the plant object is dyed / embedded so as to become reddish purple.
[0050]
Thereby, pollen and garbage are identified.
In this apparatus, one drop of the staining solution is dropped on the slide, and a cover glass on which pollen is collected by the automatic pollen collector is superimposed on this slide, and stored in a storage cassette. The driving board and IO board are the same as the automatic pollen collector and drive a total of six stepping motors.
[0051]
This device is composed of eight operating parts.
[0052]
Each unit was operated eight times and the drive accuracy was examined. As a result, the slide glass feeding device, the overlaying device, and the storage feeding device were all normal. However, although the positional deviation of about 1 mm was observed once or twice in the cover glass feeding device and the holder feeding device, this error does not affect the function. However, the most serious problem is a portion where an accurate amount of viscous dye solution is dropped onto the slide glass sent to the dyeing position in the dyeing machine. The amount of dripping was three times out of eight times.
[0053]
However, no external loss of embedded pollen was observed, and the device could be used without any problem.
[0054]
Further, the time required for processing the four samples can be completed in less than 40% of the time required by an expert to perform manually.
[0055]
(c) Data acquisition stage It is necessary to acquire pollen information necessary for image analysis with a microscope from a sample fixed on a slide glass stained with a stainer. For this purpose, this device sends the slide glass from the storage case onto the microscope stage, or automatically moves the stage with a mechanism for exchanging, and takes in data within a certain area from several predetermined positions on the slide. It has a mechanism.
[0056]
FIG. 6 is a system configuration diagram.
[0057]
The sample exchange device is basically not much different from that of the collection and staining device, but has a mechanism to do it between the stage and the objective lens.
[0058]
The following specifications were set for the control for data acquisition.
[0059]
(1) A 1 × 1 mm frame of visual field reticulated slide glass.
The relationship between the field width b and the magnification X is b = 10 ⁵ / X (μm)
[0060]
(2) Scanning range 18mm circle according to the size of the cover glass when collecting pollen [0061]
(3) Moves back and forth within the scanning range at regular intervals from the scanning pattern and one step length home position, and finally returns to the home position. The length of one step is the same as the visual field width.
[0062]
(4) Three types of scanning magnifications of 100, 400, and 1000 times.
[0063]
An XY stage control program was created based on these control specifications. The flowchart is shown in FIG.
[0064]
The experimental results are shown in FIG.
[0065]
The number of survey points in the figure indicates the number of image capture fields during the reciprocation of the 18 mm diameter scanning range from the home position and finally returning to the original position.
[0066]
As the magnification increases (the field of view is narrow), the number of survey points increases rapidly with the square of the magnification, and the pollen capture rate defined in FIG. 12 increases with the number of pollen that is divided by the boundary of the field of view as shown. It will decrease because of it.
[0067]
Also, the processing time becomes a huge value.
[0068]
On the other hand, counting and recognition of pollen by an expert is performed with an accuracy of 90% or more in about one hour for each slide. In order to speed up the processing time with the same recognition rate as this, it is necessary to grasp the position of pollen at a low magnification (100 times), eliminate unnecessary movement of the XY stage, and reduce the number of investigation points.
[0069]
Then, by improving the program, such as identifying the survey point where pollen is present at a high recognition rate of 400 times, the pollen recognition rate can be increased and the processing time can be shortened.
[0070]
(d) Image analysis device As shown in FIG. 8, the system used for image analysis is composed of a personal computer (PC-9801DS), an image analysis device (HRU-TAICHI IV-80), and an image input monochrome camera.
[0071]
The image analyzer observes the number and type of pollen on the preparation.
[0072]
Pollen images are input to a computer through a microscope, and the appearance characteristics of each pollen and the size of each pollen are examined to determine which plant pollen is attached.
[0073]
When pollen is identified manually, the magnification is 1000 times, and the counting operation is performed at a magnification of 100 to 400 times.
[0074]
For this reason, the input image is taken into the image analysis apparatus at a high magnification, and then the image analysis apparatus appropriately performs a reduction process so that the actual magnification when the image is input is set to 400 times.
[0075]
The image was shrunk / expanded to remove noise and smoothed with a median filter, and then binarized with discriminant analysis.
[0076]
For multiple pollen, labeling and fillet diameter were calculated and separated into single pollen.
[0077]
The following various values were obtained from the binarized image.
[0078]
(1) Figure maximum length, (2) Maximum vertical width, (3) Ratio of (1) and (2), (4) Figure maximum length and figure minimum width, (5) Complexity.
[0079]
On the other hand, the order of identification was performed in the order as shown in FIG. 9 including cedar having the highest ratio as a cause of pollinosis.
[0080]
In accordance with the identification sequence of FIG. 9, the values of (1) maximum figure length to (5) complexity described above were obtained, and first, the range of values that could be recognized as cedar was defined for the conditions (1) to (3). In accordance with the recognition rate of experts, the recognition rate was set to fall within 70% as a standard.
[0081]
The result is shown in FIG.
[0082]
The recognition rate in the figure is the ratio recognized as cedar, and other 1 indicates other than cedar and other 2 indicates other than cedar, macomo, and pine.
[0083]
The setting value for the maximum figure length in (1) is effective for distinguishing between cedar and cedar, and 50% of Makomo falls within the cedar setting.
[0084]
The value of the maximum vertical width in (2) indicates that no other pollen falls within the set value of cedar, but 25% of the cedar has deviated from the set value.
[0085]
In addition, the ratio of the maximum figure length to the maximum vertical width in (3) is a sufficient value for the purpose of separating cedar, makomo and pine.
[0086]
Further separation is made for those that do not meet the cedar conditions in the first treatment described above.
[0087]
Assuming that the condition of (3) is not considered, (4) the ratio of the maximum figure length to the minimum figure diameter width and (5) complexity were added.
[0088]
As a result, as shown in FIG. 14, (4) can be used separately for cedar and makomo, and cedar has less variation in value than makomo than (5), and the set value can be narrowed down.
[0089]
Next, the pollen and pine other than those identified as cedar are identified. Setting values were defined for (1) to (3).
[0090]
As a result, as shown in FIG. 15, a superior difference clearly appeared for any of the three conditions.
[0091]
The last remaining pollen was identified by Makomo.
[0092]
As a result, as shown in FIG. 16, under the condition (3), there are many variations in the values that Makomo can take, and it is difficult to specify the set values.
[0093]
However, all can be separated according to the conditions of (1) and (2).
[0094]
As described above, even when a plurality of pollen is captured as an input image at a magnification of 400 times, correct identification can be performed.
[0095]
This is also due to the fact that the size of pollen particle size on a single cover glass could be kept within a certain range by the collector.
[0096]
As described above, in order to obtain pollen type and scattering information in a short time, it is possible to develop a system that can automate pollen collection, staining, data sampling, and image analysis, and process them without skill. It was.
[0097]
【The invention's effect】
Since the present invention is configured as described above, pollen floating in the air can be automatically collected by a collection device, and this pollen is dyed so that it can be easily analyzed by a staining embedding device. The pollen that has been dyed and embedded can be measured with an image analysis device, and the type and number of pollen can be measured, and a series of operations from pollen collection to analysis can be performed automatically.
[0098]
Therefore, when pollen is collected manually and pollen is analyzed manually as in the past, there will be a difference in the pollen collection method and analysis method, leading to a situation where the data is not uniform and reliable. It is possible to solve this problem, and it is possible to perform analysis with high reliability and consistency in a short time without the need for specialized skills for collecting and analyzing pollen, and without having to manually handle pollen types and scattering information. it can.
[0099]
Further, the present invention is a different Tsubu径No pollen can be collected in a separate cover glass, can be collected with a cover glass arranged small dust behind the collecting chamber such as dust, pollen The cover glass placed in the front collection chamber where the particles are collected is free from dust and image analysis can be performed accurately.
[0100]
The present invention also transfer a capturing current pollen on a slide glass, and imbedded, in which to carry out the dyeing automatically, short and dust and pollen contained in pollen collected by collecting device It can be accurately identified without manpower for time.
[0101]
In addition, the present invention can eliminate wasteful movement of the XY stage, can accurately recognize the type and number of pollen in a short time, and can automatically count them. Therefore, it is possible to perform reliable and unified image analysis in a short period of time without requiring manual processing of pollen types and scattering information.
[Brief description of the drawings]
FIG. 1 is an explanatory view showing a configuration of a pollen collector of this example.
FIG. 2 is an explanatory diagram showing connection of a four-stage pollen collector according to the present embodiment.
FIG. 3 is an explanatory view showing a detailed structure of one stage of a pollen collector of the present example.
FIG. 4 is an explanatory diagram showing a flowchart of basic operations of the embodiment.
FIG. 5 is an explanatory diagram showing an operation control flowchart of the dyeing device of the present embodiment.
FIG. 6 is an explanatory diagram showing a configuration diagram of a data capture stage of the present embodiment.
FIG. 7 is an explanatory diagram showing a stage control flowchart of the embodiment.
FIG. 8 is an explanatory diagram illustrating a configuration diagram of an image analysis apparatus according to the present exemplary embodiment.
FIG. 9 is an explanatory diagram showing an identification order of the embodiment.
FIG. 10 is an explanatory diagram showing pollen capturing characteristics of the collector of the present example.
FIG. 11 is a table showing characteristics of the data capturing device according to the present embodiment.
FIG. 12 is an explanatory diagram showing the definition of the capture rate of the present embodiment.
FIG. 13 is a table showing a ratio (%) that falls within a set value of capture cedar.
FIG. 14 is a table showing a ratio (reprocessing) within a set value of cedar according to the present embodiment.
FIG. 15 is a table showing the ratios within the pine setting values of this example.
FIG. 16 is a table showing a ratio that falls within the setting value of Mako.
FIG. 17 is a schematic configuration explanatory diagram showing the configuration of the present embodiment in an easily understandable manner.

Claims (1)

A collecting device for automatically collecting pollen floating in the air, staining embedding sealed by staining pollen collected by the collecting device and the distinction and the pollen and dust so that the image easy analysis It consists of an embedding device and an image analysis device that measures the type and number of the pollen by analyzing the image based on the microscopic information of the pollen dyed and embedded by the dyeing and embedding device , Air is sucked into the device, passed through a series of collection chambers connected in series, and pollen in the air is collected in a cover glass placed in each chamber. The speed is configured to be variable, and the staining apparatus automatically transfers and embeds pollen collected by the collection apparatus on a slide glass, and automatically performs staining for identifying pollen. Further, the image analysis device is configured as follows: A magazine with a plurality of slide glasses with pollen stained and embedded by the staining device is installed in a microscope equipped with a CCD camera, and the slide glass is automatically inserted and ejected sequentially into the microscope XY stage. The XY stage is configured to sequentially move according to programming, transfer an image to an image analysis computer, identify the type by a programming criterion, and measure the number of pollen automatic. Collection and analysis system.

Images