JP2006349610A - Pollen particle counting system, and pollen particle imaging device, pollen particle counting device and pollen particle counting method used therefor - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a pollen particle count system, capable of imaging pollen particles using an imaging device of a simple structure and accurately counting the pollen particles, even in when pollen particles are captured in an overlapped state, and a pollen particle imaging device, a pollen particle counting device and a pollen particle counting method used for the system. <P>SOLUTION: Ultraviolet rays 44 are emitted from an ultraviolet source 11, in order to generate self-fluorescence from pollen particles captured on a slide glass 42, and the ultraviolet rays 44 transmitted by the slide glass 42 are interrupted by a short-wavelength light cut filter 13. Pollen particles which emit self-fluorescence are imaged using pixels smaller than the pollen particles, binarization processing and noise removing processing are performed thereto, based on the number of pixels of an imaging element per pollen particle area that is preset by a preprocessing means 22b of a CPU 22, and the number of pollen particles in image data is calculated by a counting means 22c of the CPU 22. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a pollen particle counting system that counts the amount of pollen scattered in the air by imaging pollen particles using autofluorescence emitted from pollen, and a pollen particle imaging device and pollen particle counter used therefor The present invention relates to a device and a pollen particle counting method.

  In recent years, the incidence of Japanese cedar pollinosis in Japan has increased to about 1 in 5 people. For this reason, during the pollen scattering season, the amount of pollen scattered is measured and predicted.

  Conventionally, the amount of pollen that has been scattered has been visually measured using a microscope after being collected. The collection method for performing the counting includes a method of collecting petrolatum or the like on a slide glass and collecting the pollen particles on the surface (Durham method), and a mesh smaller than the pollen particles. There is a method of collecting using a filter having the same.

  Non-Patent Document 1 describes a method in which pollen collected by the Durham method is dyed and then counted visually using a microscope.

  However, this method of counting visually requires a skilled person to perform the counting because pollen particles and other particles have to be distinguished and counted by humans, and it takes a lot of work time.

  In order to solve this problem, Non-Patent Document 2 describes the size of the particles by dyeing pollen collected by the Durham method and then reading the image of the pollen stained by an image scanner and performing image processing. And a technique for counting pollen identified from the shape by a computer.

  However, even in the method described in Non-Patent Document 2, there is a possibility that particles having the same size or similar shape may be erroneously recognized as pollen particles.

  In addition, there is a method of identifying and counting pollen particles by autofluorescence emitted from pollen using a fluorescence microscope, but this device can only count a limited area because it enlarges and counts the image with an optical system, There was a problem that the accuracy of the scattering amount counting result was low and the apparatus was expensive.

Furthermore, Patent Document 1 obtains an enlarged image of pollen particles that emit autofluorescence by utilizing the property that organic substances such as pollen particles emit autofluorescence when excited by ultraviolet rays. An apparatus for analyzing and counting particles is described. Only pollen particles can be automatically selected by this apparatus, and counting can be performed for each type of pollen.
All about cedar pollen (Medical Journal, authored by Sahashi, Takahashi, and Murayama) Appendix 1 How to observe airborne pollen pp.103-105 Pollen counting method with film scanner and LIA32 Ver.0.37 β1: http://www.agr.nagoya-u.ac.jp/%7Eshinkan/LIA32/p_count.html JP 11-337469 A

  However, in recent years, the resolution of image processing apparatuses has increased, and it is possible to analyze the size and shape of pollen particles without enlarging the image as in the technique of Non-Patent Document 2 or Patent Document 1. It has become to.

  In addition, pollen particles may be collected without overlapping one by one, or when the concentration is high, the particles may be collected in a state where the particles overlap each other. It does not describe the counting method when collected in an overlapping state.

  Accordingly, an object of the present invention is to provide a pollen particle counting system capable of imaging and counting pollen particles with an imaging device having a simple structure, and a pollen particle imaging device, a pollen particle counting device, and a pollen particle counting method used therefor. It is to be.

  Another object of the present invention is to provide a pollen particle counting system capable of counting with high accuracy even when pollen particles are collected in an overlapped state, and a pollen particle imaging device and pollen particle counter used therefor. An apparatus and a pollen particle counting method are provided.

  In order to solve the above problems, the pollen particle counting system of the present invention comprises an ultraviolet light source for irradiating ultraviolet light to generate autofluorescence from pollen particles collected by a preparation and a short wavelength light for blocking ultraviolet light transmitted through the preparation. The image was picked up by using an image pickup device and an image pickup device for picking up an image of pollen particles emitting auto-fluorescence by being irradiated with ultraviolet rays and having a resolution capable of representing a cut filter and one pollen particle by a plurality of pixels. A pollen particle imaging device having a recording unit for recording image data, preprocessing means for obtaining image data recorded in the imaging device and performing binarization processing and noise removal processing, and binarization processing and noise removal processing And a pollen particle counting device having a counting processing unit having counting means for counting the number of pollen particles from the image data subjected to.

  Further, the preprocessing means of the counting processing unit may have a function of performing noise removal processing using an image having a shape that does not match the shape of the pollen particles as noise.

  Further, the counting means of the counting processing unit may count the number of pollen particles in the image data based on a preset number of pixels of the image sensor per one pollen particle area.

  According to the pollen particle counting system of the present invention, and the pollen particle imaging device, the pollen particle counting device, and the pollen particle counting method used therein, autofluorescence can be obtained with an imaging device with a simple structure by using a high-resolution imaging means. It is possible to count the pollen particles.

  In addition, even when pollen particles are collected in an overlapping state, counting can be performed with high accuracy.

<Configuration of pollen counting system 1>
The configuration of the pollen counting system 1 according to the first embodiment of the present invention will be described with reference to FIG.

  The pollen counting system 1 according to the present embodiment is configured by connecting a pollen particle imaging device 10 and a pollen particle counting device 20 with a connection cable 30.

  The pollen particle imaging device 10 includes an ultraviolet light source 11, an imaging lens 12, a short wavelength light cut filter 13, an imaging element 14, a control unit 15, a recording unit 16, and an external interface 17.

  The pollen particle imaging apparatus 10 according to the present embodiment sets up a preparation that collects pollen between the ultraviolet light source 11 and the imaging lens 12 to capture an image. The prepared slide includes a sample 41 that collects pollen, a slide glass 42 on which the sample 41 is placed, and a cover glass 43 that covers the sample 41.

  The ultraviolet light source 11 directly irradiates the preparation sample 41 with ultraviolet light 44.

  The imaging lens 12 forms an optical image of the sample 41 on the image sensor 14.

  The short wavelength light cut filter 13 blocks the ultraviolet ray 44 that has passed through the preparation.

  The imaging element 14 has 5 to 6 million pixels in the present embodiment to the extent that one pollen particle can be represented by a plurality of pixels, and converts the formed optical image into an image signal.

  The control unit 15 adjusts the sensitivity and exposure time of the image sensor 14 and sends the image signal obtained by the image sensor 14 to the recording unit 16 as image data.

  The recording unit 16 records captured image data.

  The external interface 17 is an interface for connecting to the pollen particle counting device 20.

  The pollen particle counting device 20 includes a connection interface 21 and a CPU 22 as a counting processing unit.

  The connection interface 21 is an interface for connecting to the pollen particle imaging device 10.

  The CPU 22 performs preprocessing and input / output means 22a for inputting / outputting image data via the connection interface 21, preprocessing means 22b for performing binarization processing and noise removal processing as preprocessing on the input image data. Counting means 22c for counting pollen particles using the obtained image data.

<Operation of pollen counting system 1>
Next, the operation | movement of the pollen counting process by the pollen counting system 1 by this embodiment is demonstrated.

  In the pollen counting process according to the present embodiment, after the sample 41 on which the pollen particles are collected is photographed by the pollen particle imaging device 10, the number of pollen particles is counted by the pollen particle counting device 20 using this photographed image. Is done.

  First, in order to photograph the sample 41, the ultraviolet light 44 is irradiated from the ultraviolet light source 11 toward the preparation sample 41.

  In the sample 41, the organic substance containing pollen particles is excited by being irradiated with the ultraviolet ray 44, and the autofluorescence 45 is emitted.

  An image of the sample 41 in which the autofluorescence 45 is emitted is taken. When photographing is performed here, if the ultraviolet light 44 that is excitation light reaches the image sensor 14, the photographing of the autofluorescence 45 is hindered. Therefore, the ultraviolet light 44 that has passed through the preparation is blocked by the short wavelength light cut filter 13.

  The optical image obtained by photographing is imaged on the image sensor 14 by the imaging lens 12 and converted into an image signal. At this time, the control unit 15 adjusts the sensitivity and exposure time of the image sensor 14.

  The converted image signal is sent to the recording unit 16 and recorded as image data by the control unit 15.

  Since this pollen particle imaging device 10 is photographed in a state where the sample 41 is irradiated with excitation light and emits autofluorescence, an image in which organic substances including pollen particles are emphasized can be obtained.

  Next, an operation of counting the number of pollen particles by the pollen particle counting device 20 from the image data obtained by this photographing will be described with reference to the flowchart of FIG.

  The pollen particle counting device 20 calculates the number of pixels per one pollen particle area calculated from the resolution of the image data and the shape pattern of the pollen particles in order to perform the polling particle counting process, and pre-processing means of the CPU 22 in advance. 22b is set.

  In this embodiment, since the number of pixels of the imaging device 10 is 5 to 6 million pixels, the size of one pixel is 2 to 3 μm, and one pollen particle (about 30 μm) is about 60 pixels. Moreover, since the shape of the pollen particles is spherical, the shape pattern is a circle having a diameter of about 10 pixels.

  First, in the input / output means 22a of the CPU 22, the image data recorded in the recording unit 16 of the pollen particle imaging device 10 is acquired via the connection cable 30 and the connection interface 21 (S1).

  FIG. 3 is an example of the acquired image data, and an organic substance containing pollen particles is emphasized by autofluorescence.

  The acquired image data is binarized in the preprocessing means 22b of the CPU 22 (S2). FIG. 4 shows an example of a state in which the image data of FIG. 3 has been binarized. FIG. 5 shows a further enlarged state of the binarized image of FIG. In FIG. 5, dotted lines indicate real images (real images 52 and 54 of pollen particles, real images 56 other than pollen particles), and hatched portions indicate units for each pixel 51 by binarizing an autofluorescence image formed by photographing. (Self-fluorescent images 53 and 55 of pollen particles, and auto-fluorescent image 57 other than pollen).

  After the binarization processing is performed, noise processing is further performed in the preprocessing means 22b, and organic substance images other than pollen particles are removed.

  The noise processing will be described. First, an image having an area sufficiently smaller than 60 pixels, which is an area set as one pollen particle, here an image having an area of 30 pixels or less is regarded as noise other than pollen and is removed. (S3).

  Next, an image that does not have a circular arc with a diameter of about 10 pixels that is a shape pattern set in advance, for example, the autofluorescence image 57 of FIG. 5, is regarded as noise other than pollen particles and is removed (S4 ).

  The image data that has been subjected to these preprocessing is used, and the counting means 22c of the CPU 22 performs pollen particle counting processing.

  The counting process will be described. First, whether or not a plurality of pollen particles overlap is determined from the area of each pollen particle image. At this time, in the case of around 60 pixels, which is the number of pixels per one pollen particle area of the pollen particle image set in advance as a determination target, as shown in the autofluorescence image 53 of FIG. It is judged and the number of the pollen particle images is counted (S5).

  Next, when the area of the pollen particle image to be determined is sufficiently larger than 60 pixels, it is determined that a plurality of pollen particles are overlapped as shown in the autofluorescence image 55 of FIG. A value obtained by dividing the area of the particle image by 60 pixels is obtained.

  The value calculated by dividing the area of this pollen particle image by 60 pixels is smaller than the actual number due to the overlap, so the integer value rounded up after the decimal point is regarded as the number of pollen particle images. It is. For example, when the calculated value is 1 or more and less than 2, it is determined that there are two, and when it is 2 or more and less than 3, it is determined that there are three.

  In FIG. 5, the area of the autofluorescence image 55 is 171 pixels, and the value obtained by dividing this by 60 pixels is 2.85. Therefore, it is determined that the number of pollen particle images is three by rounding up the decimal part.

  In this way, the number of pollen particle images is calculated from the autofluorescence images determined to be in an overlapping state in the image data, and counted (S6).

  FIG. 6 is a graph showing the counting result in step S6. The horizontal axis indicates the area of the autofluorescence image in the image data when the area for one pollen particle is 100, and the vertical axis indicates the frequency which is the number of autofluorescence images of the corresponding area.

  In the graph, the leftmost bar graph represents the frequency of the self-fluorescent image (area about 100) of the pollen particles that do not overlap, and the bar graph on the right is an autofluorescent image (area about 200) where two pollen particles overlap. Represents the frequency of.

  The total number of pollen particles in the image data is obtained by calculating the sum of values obtained by multiplying the number of pollen particles overlapping the frequency of each bar bluff (S7).

  According to the pollen counting system of the present embodiment described above, pollen particles are counted with an imaging device having a simple structure in order to count pollen particles based on image data captured using an image sensor sufficiently smaller than the pollen particles. Can do.

  In addition, even when pollen particles are collected in an overlapping state, counting can be performed with high accuracy.

1 is an overall view showing a pollen particle counting system according to an embodiment of the present invention. It is a flowchart which shows operation | movement of the pollen particle counting apparatus used for the pollen particle counting system by one Embodiment of this invention. It is an example of the imaging data imaged with the pollen particle imaging device of the pollen particle counting system by one Embodiment of this invention. It is an example of the data which binarized the imaging data imaged with the pollen particle imaging device of the pollen particle counting system by one Embodiment of this invention. It is an enlarged view of the data which binarized the imaging data imaged with the pollen particle imaging device of the pollen particle counting system by one Embodiment of this invention. It is a graph which shows the counting result in the pollen particle counting system by one Embodiment of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Pollen counting system 10 ... Imaging device 10 ... Pollen particle imaging device 11 ... Ultraviolet light source 12 ... Imaging lens 13 ... Short wavelength light cut filter 14 ... Imaging element 15 ... Control part 16 ... Recording part 17 ... External interface 20 ... Pollen particle counting device 21 ... Connection interface 22 ... CPU
23 ... Memory 22 ... CPU
22a ... Input / output means 22b ... Pre-processing means 22c ... Counting means 30 ... Connection cable 41 ... Sample 42 ... Slide glass 43 ... Cover glass 44 ... Ultraviolet light 45 ... Autofluorescence 51 ... Pixels 52, 54, 56 ... Real images 53, 55 , 57 ... Autofluorescence image

Claims (10)

An ultraviolet light source that emits ultraviolet light to generate autofluorescence from pollen particles collected by the preparation, a short-wavelength light cut filter that blocks the ultraviolet light that has passed through the preparation, and one pollen particle are represented by a plurality of pixels. And an image sensor for imaging pollen particles that emit autofluorescence when irradiated with the ultraviolet rays, and a recording unit that records image data captured using the image sensor Pollen particle imaging device;
Preprocessing means for obtaining image data recorded in the imaging device and performing binarization processing and noise removal processing, and counting the number of pollen particles from the image data subjected to the binarization processing and noise removal processing A pollen particle counting device having a counting processing unit having a counting means;
A pollen particle counting system comprising: The pollen particle counting system according to claim 1, wherein the preprocessing means of the counting processing unit has a function of performing noise removal processing using an image having a shape that does not match the shape of the pollen particles as noise. The counting means of the counting processing unit counts the number of pollen particles in the image data based on a preset number of pixels of the image sensor per one pollen particle area. The described pollen particle counting system. An ultraviolet light source that emits ultraviolet light to generate autofluorescence from pollen particles collected in the preparation;
A short-wavelength light cut filter that blocks the ultraviolet light transmitted through the preparation;
An imaging device having a resolution capable of representing one pollen particle by a plurality of pixels and imaging pollen particles emitting autofluorescence;
A recording unit that acquires and records image data captured by the image sensor;
A pollen particle imaging apparatus comprising: Pre-processing means for performing binarization processing and noise removal processing on image data of pollen particles that have emitted autofluorescence due to ultraviolet excitation, and an image that has been subjected to the binarization processing and noise removal processing. A pollen particle counting apparatus comprising: a counting processing unit having counting means for counting the number of pollen particles from data. The pollen particle counting apparatus according to claim 5, wherein the preprocessing means of the counting processing unit has a function of performing noise removal processing using an image having a shape that does not match the shape of the pollen particles as noise. The counting means of the counting processing unit counts the number of pollen particles in the image data based on a preset number of pixels of the image sensor per one pollen particle area. The pollen particle counter described. An ultraviolet irradiation step of irradiating ultraviolet rays to generate autofluorescence from pollen particles collected in the preparation;
An ultraviolet blocking step for blocking the ultraviolet light transmitted through the preparation;
An imaging step of imaging pollen particles emitting autofluorescence using an imaging device having a resolution capable of representing one pollen particle by a plurality of pixels;
A counting process step of acquiring captured image data, performing binarization processing and noise removal processing, and counting the number of pollen particles from the image data subjected to the binarization processing and noise removal processing;
A pollen particle counting method comprising: The pollen particle counting method according to claim 8, wherein the counting process step executes a noise removal process using an image having a shape that does not match the shape of the pollen particle as noise. The polling particle counting method according to claim 8 or 9, wherein the counting processing step calculates the number of pollen particles in the image data based on a preset number of pixels per one pollen particle area. .