JP2003139678A - Cell image analyzer, cell image analyzing method and recording medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a cell image analyzer, a cell image analyzing method and a recording medium in which a cell image can be analyzed efficiently with high reliability. SOLUTION: The cell image analyzer comprises means 26 for processing the image data of a cell image obtained by imaging the cell of an animal or a plant to produce a binarized image data, means 23 for frequency converting the image data to divide the band into a frequency component on the high frequency side and a frequency component on the low frequency side including the DC component, a cell block extracting means 25 generating intermediate data for feature extraction using the frequency component at least on the high frequency side, and an analyzing means 27 for comparing the intermediate data with the binarized image data and extracting the cell part and the protruding part of an animal or a plant.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a cell image analyzing apparatus and cell image analyzing method for performing a predetermined analysis by frequency-converting image data of cell images of animals and plants, and a recording medium for executing the method. Is.

[0002]

2. Description of the Related Art In various tests in the field of biochemistry, cells and microorganisms of animals and plants are cultured under various conditions, and the changes and the degree of growth over time are observed. For example, in the case of observing the sequentially growing portions of protrusions such as nerve cells, these are tracked over time and observed.
It is necessary to record and convert it into data. Conventionally, this kind of work relies solely on human labor, and the person in charge of the experiment visually observes the cells within the microscope's field of view, pays attention to the observation target site such as the neurite, and creates and records the necessary items. I was working.

[0003]

However, such an observation work is a troublesome work that requires a lot of labor and time, and this observation causes a decrease in the efficiency of the entire experiment work, and also causes an excessive amount of labor to the person in charge of the experiment. Had the problem of compelling the work load of. In addition, when the observation results are converted into data, there is a problem that it is difficult to ensure the reliability of the data because the results converted into data are likely to vary depending on the experience and skill of each person in charge of observation. It was

In this cell image analysis device, cell image analysis method and recording medium, it is required to analyze cell images efficiently and with high reliability.

In order to meet this requirement, the present invention provides a cell image analyzer capable of efficiently analyzing cell images with high reliability, and a cell image analyzing apparatus for efficiently analyzing cell images with high reliability. An object of the present invention is to provide an image analysis method and a recording medium for executing the method.

[0006]

In order to solve the above-mentioned problems, a cell image analyzer of the present invention binarizes image data of a cell image obtained by imaging cells of animals and plants to binarize the image data. , A frequency conversion means for frequency-converting the image data and band-dividing the frequency component into a high-frequency side frequency component and a low-frequency side frequency component including a direct current component, and at least a high-frequency side It has a cell block extraction means for creating intermediate data for feature extraction using frequency components, and an analysis processing means for comparing the intermediate data with the binarized image data to extract the cell parts and protrusions of animals and plants. It has a configuration.

As a result, it is possible to obtain a cell image analyzer capable of efficiently analyzing a cell image with high reliability.

In order to solve the above problems, the cell image analysis method of the present invention is a binarization in which image data of a cell image obtained by imaging cells of animals and plants is binarized to generate binarized image data. A processing step, a frequency conversion step of frequency-converting the image data to divide the frequency component into a high frequency side frequency component and a low frequency side frequency component including a direct current component, and a characteristic using at least the high frequency side frequency component It is provided with a configuration including a cell block extraction step of creating intermediate data for extraction, and an analysis processing step of comparing the intermediate data with the binarized image data to extract cell parts and projections of animals and plants.

As a result, a cell image analysis method for efficiently analyzing cell images with high reliability can be obtained.

In order to solve the above problems, a recording medium of the present invention has a structure in which a program for executing each step of the above method is recorded.

As a result, a recording medium for carrying out the above method can be obtained.

[0012]

BEST MODE FOR CARRYING OUT THE INVENTION The cell image analyzer according to claim 1 of the present invention binarizes image data of a cell image obtained by imaging cells of animals and plants to generate binarized image data. Binarization processing means, frequency conversion means for frequency-converting the image data to divide the frequency components into high-frequency side frequency components and low-frequency side frequency components including DC components, and at least high-frequency side frequency components. It has a cell block extracting means for creating intermediate data for feature extraction by using it, and an analysis processing means for comparing the intermediate data with the binarized image data to extract the cell parts and protrusions of animals and plants. It is a thing.

With this configuration, it is possible to compare the intermediate data for feature extraction with the binarized image data and extract the cell portion and the protrusion portion of the animal or plant without manual labor, so that the cell image can be efficiently generated. In addition, it has an effect that analysis can be performed with high reliability.

The cell image analyzer according to a second aspect is the cell image analyzer according to the first aspect, wherein the cell block extracting means uses one or more frequency components of two frequency components. The frequency domain of the frequency component to be divided is divided into a plurality of blocks, and the cell block is extracted from the plurality of blocks.

With this configuration, the binarized image data can be divided into a plurality of blocks to accurately determine the block content of the cell block or the projection block.

A cell image analyzer according to a third aspect is the cell image analyzer according to the second aspect, wherein the cell block extracting means utilizes the directional strength of the blocks and the arrangement pattern to project the block from the cell blocks. Is to be excluded.

With this configuration, since the protrusion block can be accurately determined, the protrusion block can be accurately removed from the cell block.

A cell image analyzer according to a fourth aspect is the cell image analyzer according to the second aspect, wherein the analysis processing means surrounds the cell block when the cell block and the binarized image data are superposed. If there is binarized image data, it is judged whether the cell block is a cell part, and if there is no binarized image data, it is judged that it is a medium, and fitting of cells on the binarized image data is performed. By doing so, the cell portion is detected from the cells having the protrusions.

With this configuration, it is possible to accurately determine whether a cell portion, a protrusion portion, or a culture medium. Therefore, by fitting a cell on the binarized image data, a cell portion can be accurately extracted from a cell having a protrusion portion. It has the effect that it can be extracted.

The cell image analyzer according to a fifth aspect is the cell image analyzer according to the fourth aspect, wherein the analysis processing means has an average value of the entire image or an average value of DC components of frequency components on the low frequency side. And the average value of the neighborhood of the target block are compared to determine whether the target block is the cell part or the medium.
The threshold value of the binarized image data is made variable.

With this configuration, the threshold value of the binarized image data can be accurately set, and the binarized image data can be accurately generated from the original image data.

According to a sixth aspect of the present invention, there is provided a cell image analysis method, which comprises a binarization processing step of binarizing image data of a cell image obtained by imaging cells of animals and plants to generate binarized image data. A frequency conversion step of frequency-converting the image data to divide the frequency component into a high frequency side frequency component and a low frequency side frequency component including a direct current component, and an intermediate for feature extraction using at least the high frequency side frequency component. It has a cell block extraction step of creating data and an analysis processing step of comparing the intermediate data and the binarized image data to extract the cell parts and projections of the animal and plant.

With this configuration, it is possible to compare the intermediate data for feature extraction with the binarized image data and extract the cell portion and the protrusion portion of the animal or plant without manual labor, so that the cell image can be efficiently obtained. In addition, it has an effect that analysis can be performed with high reliability.

The cell image analyzing method according to claim 7 is the cell image analyzing method according to claim 6, wherein one or more frequency components of two frequency components are used in the cell block extracting step. The frequency domain of the frequency component to be divided is divided into a plurality of blocks, and the cell block is extracted from the plurality of blocks.

With this configuration, the binarized image data can be divided into a plurality of blocks to accurately determine the block content of the cell block or the protrusion block.

The cell image analyzing method according to claim 8 is the cell image analyzing method according to claim 7, wherein the cell block extracting means utilizes the directional strength of the blocks or the arrangement pattern to project the block from the cell block. Is to be excluded.

With this configuration, since the protrusion block can be accurately determined, the protrusion block can be accurately removed from the cell block.

A cell image analysis method according to a ninth aspect is the cell image analysis method according to the seventh aspect, wherein in the analysis processing step, when the cell block and the binarized image data are overlapped, the cell block is surrounded by the cell block. If there is binary image data, it is judged whether the cell block is the cell part,
If there is no binarized image data, it is judged as a medium,
By fitting the cells on the binarized image data, the cell parts are detected from the cells having the protrusions.

With this configuration, it is possible to accurately determine whether it is a cell portion, a protrusion portion, or a culture medium. Therefore, by fitting a cell on the binarized image data, a cell portion can be accurately extracted from a cell having a protrusion portion. It has the effect that it can be extracted.

The cell image analysis method according to claim 10 is:
10. The cell image analysis method according to claim 9, wherein the analysis processing means compares the average value of the entire image or the average value of the DC component of the frequency components on the low frequency side with the neighborhood average value of the target block, The threshold value of the binarized image data is made variable by determining whether the target block is the cell part or the culture medium.

With this configuration, the threshold value of the binarized image data can be accurately set, and the binarized image data can be accurately generated from the original image data.

The recording medium according to claim 11 is the recording medium according to claim 6.
The program for executing each step of the cell image analysis method according to any one of 1 to 10 is recorded.

With this configuration, there is an effect that the cell image analysis method according to any one of claims 6 to 10 can be executed at any place at any time.

FIG. 1 shows an embodiment of the present invention.
~ It demonstrates using FIG.

(Embodiment 1) FIG. 1 is a block diagram showing a cell image analyzer according to Embodiment 1 of the present invention.

In FIG. 1, 1 is a positioning stage for positioning the container 2, 3 is a sample housed in the container 2, and 4 is a sample.
Is a microscope for viewing cells, 5 is a camera that outputs captured images as digital image data, 6 is an illuminating device that illuminates the container 2, 10 is a mechanism control unit that controls the positioning stage 1 and the microscope 4, and 11 is a camera 5. An image storage unit that stores digital image data from the device, 12 is a first storage unit that stores a program, 13 is a storage medium drive that drives a storage medium, and 14a, 14b, and 14c are first, second, and third storage units. A storage medium, a central processing unit 15 as a CPU, a display unit 16 as a monitor device, and a second storage unit 17 for storing data.
The storage unit 18, an operation / input unit such as a keyboard and a mouse for inputting operation commands and data, and a communication unit 19 for exchanging data with an external device via a wired or wireless communication network.

The configuration and function of the cell image analysis apparatus thus configured will be described.

In FIG. 1, a container 2 which is a carrier for accommodating cells to be observed is placed on the positioning stage 1. The container 2 is a transparent body that is visible from the outside, and the sample 3 containing the cells of the animal or plant to be analyzed is stored inside. A microscope 4 having a focus mechanism is arranged below the positioning stage 1, and a camera 5 is arranged below the microscope 4.

With the illumination device 6 arranged above the positioning stage 1 turned on to illuminate the container 2, the microscope 4
The camera 5 acquires the digital image data of the cell image by imaging the cells of the animal and plant in the sample 3 in the container 2 via the camera 5. The positioning stage 1 and the microscope 4 are connected to the mechanism control unit 10. By controlling the positioning stage 1 by the mechanism control unit 10, a desired position in the container 2 can be imaged by the camera 5, and the microscope can be used. By controlling the focus mechanism of No. 4, it is possible to focus on a desired position.

The camera 5 is connected to the image storage unit 11, and the digital image data acquired by the camera 5 is stored in the image storage unit 11. The first storage unit 12 performs various calculations such as a program for performing frequency conversion processing and a program for performing analysis processing based on the converted data.
A program that performs processing of operation control is stored. The central processing unit 15 executes various calculations and operation controls according to the programs stored in the first storage unit 12. Display unit 16
Displays an image acquired by the camera 5 and an image subjected to analysis processing, and also displays a guidance screen for operation / input.

The second storage unit 17 stores various data such as analysis results, various processed data, and setting data. The storage medium drive 13 is a drive device that reads data from a portable storage medium such as a magnetic disk such as a floppy (registered trademark) disk or a memory card. Here, different types of data are read from the following three types of storage media via the storage medium drive 13 as needed.

The first storage medium 14a stores the image data of the original image before the analysis processing, and carries out the image obtained by a certain cell image analyzer or other cell images. The image acquired by the analyzer can be analyzed, for example. The second storage medium 14b is a program for performing frequency conversion processing,
A program for performing various calculations and operation control processes, such as a program for performing analysis processing based on the converted data, is stored, and the calculation unit, the storage unit, the operation / input unit, and the display unit are provided. By installing these programs in a personal computer, a general personal computer can function as a cell image analyzer. That is, by analyzing and processing the image data read from the first storage medium 14a in the personal computer according to the program read from the second storage medium 14b, the analysis processing of the cell image can be performed. The analysis result data is stored in the third storage medium 14c. Thirdly, the already-analyzed data already created by a certain cell image analyzer is written in the third storage medium 14c and brought out, or the data of the analysis result performed by the external analyzer is written.
By reading the analysis result data from the storage medium 14c, the analysis result can be displayed on the display unit 16 as a monitor device.

Next, the processing function of the cell image analyzer will be described with reference to FIG. FIG. 2 is a functional block diagram showing processing functions of the cell image analyzer according to the first embodiment of the present invention.

In FIG. 2, the frequency conversion means 23, the cell block extraction means 25, the binarization processing means 26, the analysis processing means 27, the display processing means 28, and the recording means 29 are the same as those shown in FIG. The function realizing means is realized by executing the program stored in the storage unit 12. Further, the frequency conversion data storage unit 24 and the analysis result storage unit 30 indicate storage areas set in the second storage unit 17 shown in FIG. The camera 5, the display unit 16, and the operation / input unit 18 are the same as those in FIG.
The original image storage unit 20 corresponds to the image storage unit 11 in FIG.

In FIG. 2, the original image storage unit 20 stores the image data acquired by the camera 5 and the image data read from the first storage medium 14a. The image data to be analyzed is first frequency-converted by the frequency conversion means 23. The frequency conversion method is a method of analyzing a signal such as a voice or an image by using frequency conversion. In the process of extracting a specific waveform from various waveforms included in the original signal to be analyzed, the extraction target waveform It has a feature that the location information of the is stored without being damaged.

In FIG. 2, cell block extracting means 25
Divides the conversion signal stored in the frequency conversion data storage unit 24 into blocks and determines for each block whether it is a cell block or a cell block other than a cell block. The binarization processing means 26
Indicates in binary image data whether or not there is a target (presence or absence of an image) in order to extract the cell portion and the protrusion portion in detail.
Here, the cell block refers to a block determined to be a cell part.

Here, an example of a processed image in which frequency conversion, cell block extraction processing, and binarization processing are applied to image analysis of a cell image will be described with reference to FIGS. 3 and 4. Figure 3
(A), (b), (c), FIG. 4 (a), (b), (c)
FIG. 3 is an image diagram showing a cell image in the cell image analyzer according to the first embodiment of the present invention.

FIG. 3A shows an original image of a microscope image of animal and plant cells. This original image is frequency-converted and the converted signal 40 divided for each frequency band is divided into block regions, and there is a process for determining whether each divided block is a cell block or a block other than cells, and whether there is a target. A process of binarizing the luminance data (original image data) of the original image to generate binarized image data in order to make a fine determination as to whether or not the determined cell block and the binarized image data are compared The process of extracting a portion and a protrusion will be described with reference to FIGS. 3 and 4. In the image of FIG. 4 (b),
Image elements such as cell patterns and dendrites composed of high-frequency components appear.

In the present embodiment, in the process of obtaining the intermediate data for performing the analysis processing for extracting the characteristic portion of the cell image from the frequency-converted image signal (converted signal), for example, the characteristic is extracted from the high frequency component. Method is used. That is, here, the intermediate data is created using only the high frequency components. This process is executed by the cell block extracting means 25, as shown in FIG.
The synthesized high frequency component data (intermediate data) is stored in the frequency conversion data storage unit 24. Analysis processing means 27
Performs analysis processing based on this data.

This cell image analyzer is configured as described above. Hereinafter, a method of analyzing a cell image will be described with reference to FIG. FIG. 5 is a flowchart showing the cell image analysis method according to the first embodiment of the present invention.

In FIG. 5, first, a container 2 containing a sample 3 containing animal and plant cells to be analyzed is placed on a positioning stage 1.
The sample 3 is placed on the top, the sample 3 is imaged by the camera 5, and the image data including the cell image of the analysis target is captured in the original image storage unit 20 (S1). As a result, the original image 50 as shown in FIG. 3A is obtained.

The original image 50 has cell parts 5 of different shapes.
1, 52, and 53 are imaged, and thin line-shaped dendrites (projections) 51a extending from the cell part 51 and the cell part 52
From the dendrites 52a, 52b having different thicknesses,
Similarly, dendrites 53a and 53b extending from the cell portion 53 and having different thicknesses are displayed.

Next, the frequency conversion means 23 performs frequency conversion on the image data (S2).
The converted signal obtained by the frequency conversion is stored in the frequency conversion data storage unit 24 for each frequency band. This frequency conversion is repeated a predetermined number of times. That is, of the data stored in the frequency conversion data storage unit 24 after the conversion processing, the data in the low frequency space (DC component, frequency component on the low frequency side)
Is again read by the frequency conversion means 23, and frequency conversion is executed again.

Next, the high frequency component (frequency component on the high frequency side) obtained by the frequency conversion is stored in the high frequency component storage section of the frequency conversion data storage section 24. Next, the cell block extraction means 25 divides the high frequency component into blocks (S3). A large amount of information indicating dendrites remains in the image of the high frequency component, and the unnecessary portion indicated by the frequency component on the low frequency side is removed.

Next, the cell block extracting means 25 extracts the cell block based on the image data (S).
4). Here, from the image shown in FIG.
A process for removing 1, 52, 53 is first performed.
Here, since the cell parts 51, 52, and 53 correspond to dense parts on the image, they are divided into blocks like the image shown in FIG. 3C, and are divided into blocks as shown in FIG. As shown, only the cell parts 51, 52, 53 are separated.

Next, the binarization processing means 26 binarizes the original image as shown in FIG. 4B (S5), and the analysis processing means 2
7 shows the cell parts 51, 52, 53 from the image shown in FIG.
Is removed, the image data for analysis processing shown in FIG. 4C, that is, the dendrites 51a, 5 from the original image is removed.
The image data in which 2a and 53a are extracted is obtained. The analysis processing means 27 superimposes and analyzes the binarized image data and the cell block pattern (a pattern in which at least one or more frequency components of the converted signal are blocked, intermediate data), and thus detailed lines can be measured. Then, based on this image data for analysis processing, analysis processing means 27
Executes analysis processing (S6).

FIG. 6 is a block diagram showing the analysis processing means 27 of FIG. As shown in FIG. 6, analysis processing means 27
Has a dendrite extraction means (feature extraction means) 45, a coordinate data storage section 46, a length storage section 47, and a branch number storage section 48.

The dendrite extracting means 45 performs image processing on the binarized image data to detect and extract dendrites of cells as the characteristic portion of the analysis target from the binarized image data. In this detection process, the coordinate data indicating the position of each dendrite, the data indicating the length of the dendrite, and the number of branches appearing in the linear portion such as the dendrite are detected. Then, the data (position data) regarding the position coordinates that specify the detected position of the dendrite in the original image is stored in the coordinate data storage unit 46. By using the position data, when the analysis result is displayed on the monitor of the display unit 16, the extracted tree protrusions are displayed and clarified with a specific color or line type and then superimposed on the original image. The analysis result can be easily compared and compared with the original image at the time of analysis.

The display processing means 28 displays the analysis result of the analysis processing means 27 and the cell image stored in the original image storage section 20 in accordance with the operation from the operation / input section 18 and the display section 16
To display. When a command to display the position of the dendrite on the original image is issued from the operation / input unit 18, the display unit 16
In addition to displaying the original image, the dots and lines indicating the positions of the dendrites are displayed on the displayed original image based on the coordinate data read from the coordinate data storage unit 46. The display processing unit 28 and the display unit 16 are display units that superimpose and display an image based on the dendrite position data on the cell image.

The recording means 29 performs a process of storing various data in the analysis result storage section 30 in response to an instruction input from the operation / input section 18. The data to be recorded includes the original image data (cell image) stored in the original image storage unit 20, frequency-converted frequency-converted data (converted signal), and analysis results analyzed by the analysis processing means 27 ( Dendritic position data and numerical data) are included. Then, when the analysis result is stored in the analysis result storage unit 30, at least the analysis result, the original image data (or the compressed image data), and the microscope magnification at the time of capturing the image are stored in association with each other. By associating such data, when analyzing the analysis results,
It is possible to compare the obtained compressed image and dendrite position data and numerical data by superimposing them on the monitor device of the display unit 16 in a state where the magnification and relative position are matched with the original image itself. Becomes As a result, the analysis processing can be performed more efficiently and precisely and finely.

Further, when extracting dendrites from a cell image, numerical data indicating the length of dendrites and the number of branches of dendrites are also detected and output, so that the conventional method is performed only by visual observation. It is possible to automate and quantify the observation work of cell images, data collection, and recording work, which improves the efficiency of work and eliminates the variation of data due to individual differences of experiment personnel, and the reliability of analysis results Can be secured. Furthermore, in the extraction of the dendrites,
By obtaining and storing the dendrite position data in the original image, it is possible to superimpose and display the dendrites extracted based on the position data on the cell image when analyzing the analysis result, Analysis work can be performed efficiently. Furthermore, by recording a program for carrying out each of the above-mentioned processes in a portable recording medium, a general personal computer can read these programs and cell image data from this recording medium, and a cell image analyzer. Can function as. That is, each of the above processes can be performed at any place and at any time.

As described above, according to this embodiment, the image data of the cell image obtained by imaging the cells of plants and animals
Binarization processing means 2 for binarizing and generating binarized image data
6, frequency conversion means 23 for frequency-converting the image data to divide the frequency component into a high frequency side frequency component and a low frequency side frequency component including a direct current component, and at least a high frequency side frequency component. Cell block extraction means 25 for creating intermediate data for extraction, and intermediate data and 2
By having the analysis processing means 27 for comparing the binarized image data with the binarized image data by comparing with the binarized image data by comparing the binarized image data with the cell parts of the plants and animals. Since it is possible to extract the cell portion and the protrusion portion of the animal and plant, it is possible to analyze the cell image efficiently and with high reliability.

Further, the cell block extraction means 25 uses one or more frequency components of the two frequency components to divide the frequency region of the frequency components to be used into a plurality of blocks and to extract the cell blocks from the plurality of blocks. By extracting, it is possible to divide the binarized image data into a plurality of blocks and accurately determine the block content of a cell block or a protrusion block.

(Second Embodiment) The configuration of the cell image analyzer according to the second embodiment of the present invention is the same as that of the first embodiment, as shown in FIGS. 1, 2 and 6.

With respect to the cell image analyzer having the above-mentioned structure, a method for determining a protrusion block will be described with reference to FIG. FIG. 7 is an image diagram showing a cell image in the cell image analyzer according to the second embodiment of the present invention. here,
The protrusion block means a block having a protrusion.

As a projection block determination method, when extracting a cell block in the frequency space, a method other than the cell block, that is, a projection block is determined by observing the intensity of the directional component in the block and the shape of the connection of the blocks. explain. Note that this determination is performed by the cell block extraction means 25.

FIG. 7 shows a state in which the image shown in FIG. 3B is divided into blocks. Normally, when determining a cell block, in each corresponding block, it is determined as a cell block by the threshold value or the number of thresholds in the block or more, but the block has directionality and the shape of the connection of the adjacent block is determined. Is determined to be a connection of protrusion-shaped blocks, the block is determined not to be a cell block but a protrusion block. For example, when looking at the protrusion 52b in FIG. 7, when the block b1 is a cell block and the directionality of the horizontal component is strong, two cell blocks b2 and b3 are connected next to each other to connect the cell blocks b2 and b3 to the upper and lower sides. If the block is not a cell block, this block makes a mistake by mistakenly identifying the thick laterally extending protrusion as a cell portion.Therefore, adding a process that removes it from the cell block as a protrusion block prevents incorrect extraction of the block. Since the number is reduced and the determination is made for each block, the processing can be performed at high speed.

As described above, according to the present embodiment, the cell block extracting means 25 accurately determines the protrusion block by removing the protrusion block from the cell block by utilizing the directional strength of the block and the arrangement pattern. Therefore, the protrusion block can be accurately removed from the cell block.

(Third Embodiment) The configuration of the cell image analyzer according to the third embodiment of the present invention is the same as that of the first embodiment as shown in FIGS. 1, 2 and 6.

The cell fitting of the cell image analysis apparatus thus constructed will be described with reference to FIG. 8A and 8B are image diagrams showing a cell image in the cell image analyzer according to the third embodiment of the present invention. The cell fitting is performed by the analysis processing means 27.
Done in.

As shown in FIG. 8 (a), when a cell is extracted with a block, the actual cell shape and the extracted block portion deviate from each other. In addition, since the protrusions are thin, it is necessary to perform analysis using the binary image data of the original image, and therefore it is necessary to fit the actual cell shape. Properly arranging the extracted cell parts and protrusions so as to correspond to the actual cell shape is called cell fitting. Cell fitting will be described with reference to FIG.

When the cell block of the frequency space data is superimposed on the binarized image of the original image for cell fitting, the block corresponding to the cell block surrounded by the cell blocks, such as the block b1 of the cell part 53, is binarized. And "0"
If it becomes "1", the cavity inside the cell can be filled. When the binarized image data continues around the cell block like the block b2 and the block b3 of the cell part 53, it is determined to be the cell part 53 even if it is not a cell block, and fitting is performed more accurately. The cell portion 53 can be determined. In this way, it is possible to more accurately determine the cell portion 53, as in 51A, 52A, and 53A in FIG. 8B.

As described above, according to the present embodiment, the analysis processing means 27 causes the cell block when there is binary image data around the cell block when the cell block and the binary image data are overlapped. Is a cell portion 53. If there is no binarized image data, it is determined to be a medium, and by fitting the cells on the binarized image data, cells having protrusions are converted into cells. By detecting the portion 53, it is possible to accurately determine whether it is a cell portion, a protrusion portion, or a culture medium. Therefore, by fitting a cell on the binarized image data, cells having a protrusion portion can be accurately detected. 53 can be extracted. In addition, by recording the program for performing the cell fitting in a portable recording medium, the program and the cell image data are read from the recording medium into a general personal computer and are made to function as a cell analyzer. be able to. That is, the fitting of the cells can be performed at any place and at any time.

(Embodiment 4) The configuration of the cell image analyzer according to Embodiment 3 of the present invention is the same as that of Embodiment 1, as shown in FIGS. 1, 2 and 6.

A method of preventing cavities and preventing dust of the cell image analyzer having the above-described structure will be described with reference to FIG. 9A, 9B, and 9C are image diagrams showing cell images in the cell image analyzer according to the fourth embodiment of the present invention. The above-mentioned processing for preventing cavities and dust is carried out by the analysis processing means 27.

When the cell image is binarized, the binarization is performed using a dynamic binarization method in which the marginal mean value of the target data is obtained and the marginal mean value is used as a comparison value of the binarization. Even when going, when the inside of the cell is brighter than the surroundings of the cell as shown in FIG. 9A, the inside of the cell is hollow as shown in a ′ of FIG. 9B. Further, since the dust of the medium is also darker than that of the medium in the dynamic binarization, the dust appeared as binarized image data as shown in b ′ of FIG. 9B. Therefore, with the average value of the entire original image or the average value of the low frequency components as the reference value, the peripheral average value of the cell part takes a value lower than the average value of the whole, so if the peripheral average value is lower than the reference value,
Bright data can also be binarized to set a bit and blacken it to prevent cavities. Also, since the peripheral average value of the medium takes a value higher than the overall average value even if dust is mixed, if the peripheral average value is higher than the reference value, dark data should also be binarized and turned white without setting bits. , Noise due to dust can be removed (see FIG. 9C). Further, when the average value of the low frequency components is used as the reference value, the number of calculations is reduced as compared with the calculation of the average value of the original image, so that the processing can be performed at high speed.

As described above, according to the present embodiment, the analysis processing means 27 compares the average value of the entire image or the average value of the DC component of the frequency components on the low frequency side with the neighborhood average value of the target block. The threshold value of the binarized image data is accurately set by determining whether the target block is the cell part or the medium and making the threshold value of the binarized image data variable, and binarized from the original image data. Image data can be accurately generated. In addition, by recording a program for carrying out the above cavity prevention and dust prevention processing in a portable recording medium, the above program and cell image data can be read from this recording medium into a general personal computer for cell analysis. It can function as a device. That is, the above-mentioned cavity prevention and dust prevention processing can be performed at any place and at any time.

[0078]

As described above, according to the cell image analyzer of the first aspect of the present invention, the image data of the cell image obtained by imaging the cells of the animal and plant is binarized. Binarization processing means for generating image data, frequency conversion means for frequency-converting the image data to divide the frequency component into frequency components on the high frequency side and frequency components on the low frequency side including the DC component, and at least high frequency Cell block extraction means for creating intermediate data for feature extraction using the frequency component on the side, and analysis processing means for comparing the intermediate data with the binarized image data to extract the cell parts and projections of animals and plants. With the above, since it is possible to compare the intermediate data for feature extraction with the binarized image data and extract the cell portion and the protrusion portion of the animal and plant without human intervention, the cell image can be efficiently produced. Advantageous effect can be analyzed in the stomach reliability.

According to the cell image analyzer of claim 2, in the cell image analyzer of claim 1, the cell block extracting means uses one or more frequency components of the two frequency components. , The frequency domain of the frequency component to be used is divided into a plurality of blocks, and the cell block is extracted from the plurality of blocks, so that the binarized image data is divided into a plurality of blocks to accurately determine whether the block is a cell block or a protrusion block. The advantageous effect that the block content can be determined is obtained.

According to the cell image analyzing apparatus of the third aspect, in the cell image analyzing apparatus of the second aspect, the cell block extracting means extracts the cell blocks from the cell blocks by utilizing the directional strength of the blocks and the arrangement pattern. By removing the protrusion block, the protrusion block can be accurately determined, so that there is an advantageous effect that the protrusion block can be accurately removed from the cell block.

According to the cell image analyzing apparatus of the fourth aspect, in the cell image analyzing apparatus of the second aspect, the analysis processing means is configured to detect the cell block when the cell block and the binarized image data are superposed. If there are binary image data around, it is judged whether the cell block is a cell part,
If there is no binarized image data, it is judged as a medium,
By performing cell fitting on the binarized image data to detect the cell portion from the cells having the protrusion portion, it is possible to accurately determine whether it is the cell portion, the protrusion portion, or the culture medium. The advantageous effect that the cell part can be accurately extracted from the cell having the protrusion by performing the cell fitting on the image data is obtained.

According to the cell image analyzer of the fifth aspect, in the cell image analyzer of the fourth aspect, the analysis processing means includes an average value of the entire image and a DC component of a frequency component on the low frequency side. By comparing the average value with the neighborhood average value of the target block, it is determined whether the target block is the cell part or the medium, and the threshold value of the binarized image data is made variable,
The advantageous effect that the threshold value of the binarized image data can be accurately set and the binarized image data can be accurately generated from the original image data is obtained.

According to the cell image analysis method of the sixth aspect, the binarization processing step of binarizing the image data of the cell image obtained by imaging the cells of animals and plants to generate the binarized image data. And a frequency conversion step of frequency-converting the image data to divide the frequency component into a high frequency side frequency component and a low frequency side frequency component including a direct current component, and for feature extraction using at least the high frequency side frequency component. By having a cell block extraction step of creating intermediate data of and an analysis processing step of comparing the intermediate data with the binary image data to extract the cell parts and projections of animals and plants, Since it is possible to extract the cell portion and the protrusion portion of the animal and plant by comparing the intermediate data for feature extraction with the binary image data, the cell image can be efficiently and highly reliable. Advantageous effect that it is possible to analyze is obtained.

According to the cell image analyzing method of the seventh aspect, in the cell image analyzing method of the sixth aspect, in the cell block extracting step, one of two frequency components is used.
Using the above frequency components, the frequency domain of the frequency components to be used is divided into multiple blocks, and the cell blocks are extracted from the multiple blocks to divide the binary image data into multiple blocks and accurately In addition, the advantageous effect that the block contents of the cell block and the protrusion block can be determined can be obtained.

According to the cell image analyzing method of the eighth aspect, in the cell image analyzing method of the seventh aspect, the cell block extracting means uses the directional strength of the blocks and the arrangement pattern to extract the cell blocks. By removing the protrusion block, the protrusion block can be accurately determined, so that there is an advantageous effect that the protrusion block can be accurately removed from the cell block.

According to the cell image analysis method of the ninth aspect, in the cell image analysis method of the seventh aspect, in the analysis processing step, when the cell block and the binarized image data are superposed, If there is binarized image data around it, it is determined whether or not the cell block is a cell part, and if there is no binarized image data, it is determined that it is a medium, and the cell block is identified on the binarized image data. By detecting the cell part from the cells having the protrusions by performing the fitting, it is possible to accurately determine whether the cells are the protrusions or the culture medium. Therefore, the cells are fitted on the binarized image data. The advantageous effect that the cell portion can be accurately extracted from the cell having the protrusion portion can be obtained.

According to the cell image analyzing method of the tenth aspect, in the cell image analyzing method of the ninth aspect, in the analysis processing means, the average value of the entire image and the DC component of the frequency component on the low frequency side are included. Threshold value of the binarized image data by determining whether the target block is a cell part or a medium and making the threshold value of the binarized image data variable by comparing the average value of Is set accurately, and binary image data can be accurately generated from the original image data, which is an advantageous effect.

According to the eleventh aspect of the recording medium, the program for executing each step of the cell image analysis method according to any one of the sixth to tenth aspects is recorded, and the sixth to sixth aspects are recorded. An advantageous effect that the cell image analysis method according to any one of 10 can be performed at any place at any time is obtained.

[Brief description of drawings]

FIG. 1 is a block diagram showing a cell image analyzer according to a first embodiment of the present invention.

FIG. 2 is a functional block diagram showing processing functions of the cell image analyzer according to the first embodiment of the present invention.

FIG. 3A is an image diagram showing a cell image in the cell image analyzer according to the first embodiment of the present invention. FIG. 3B is an image diagram showing a cell image in the cell image analyzer according to the first embodiment of the present invention. ) An image diagram showing a cell image in the cell image analyzer according to the first embodiment of the present invention.

4A is an image diagram showing a cell image in the cell image analyzer according to the first embodiment of the present invention. FIG. 4B is an image diagram showing a cell image in the cell image analyzer according to the first embodiment of the present invention. ) An image diagram showing a cell image in the cell image analyzer according to the first embodiment of the present invention.

FIG. 5 is a flowchart showing a cell image analysis method according to the first embodiment of the present invention.

FIG. 6 is a block diagram showing the analysis processing means of FIG.

FIG. 7 is an image diagram showing a cell image in the cell image analyzer according to the second embodiment of the present invention.

8A is an image diagram showing a cell image in the cell image analyzer according to the third embodiment of the present invention. FIG. 8B is an image diagram showing a cell image in the cell image analyzer according to the third embodiment of the present invention.

9A is an image diagram showing a cell image in the cell image analyzer according to the fourth embodiment of the present invention. FIG. 9B is an image diagram showing a cell image in the cell image analyzer according to the fourth embodiment of the present invention. ) An image diagram showing a cell image in the cell image analyzer according to Embodiment 4 of the present invention

[Explanation of symbols]

1 stage 2 container 3 sample 4 microscope 5 camera 6 lighting device 10 mechanism control unit 11 image storage unit 12 first storage unit 13 storage medium drives 14a, 14b, 14c storage medium 15 central processing unit 16 display unit 17 second storage Part 18 Operation / input part 19 Communication part 20 Original image storage part 23 Frequency conversion means 24 Frequency conversion data storage part 25 Cell block extraction means 26 Binarization processing means 27 Analysis processing means 28 Display processing means 29 Storage means 30 Analysis result storage Part 45 Dendrite extracting means 46 Coordinate data storage 47 Length storage 48 Branch number storage 51, 52, 53 Cell parts 51a, 52a, 52b, 53a Dendrites (projections)

Claims (11)

[Claims] 1. A binarization processing means for binarizing image data of a cell image obtained by imaging cells of animals and plants to generate binarized image data, and frequency conversion for frequency-converting the image data. A frequency conversion means for band-dividing the frequency component on the high frequency side and the frequency component on the low frequency side including the direct current component,
Cell block extraction means for creating intermediate data for feature extraction using at least the frequency component on the high frequency side and the intermediate data and the binarized image data are compared to extract cell parts and protrusions of plants and animals. A cell image analysis apparatus, comprising: 2. The cell block extracting means uses one or more frequency components of the two frequency components to divide the frequency domain of the frequency components to be used into a plurality of blocks, and to divide the plurality of blocks from the plurality of blocks. The cell image analyzer according to claim 1, wherein the cell block is extracted. 3. The cell image analyzer according to claim 2, wherein the cell block extracting means removes the protrusion block from the cell block by utilizing the directional strength of the blocks and the arrangement pattern. 4. The analysis processing means determines whether the cell block is a cell part when the binarized image data is present around the cell block when the cell block and the binarized image data are overlapped. It is determined that the medium is a medium when there is no binarized image data, and the cell portion is detected from cells having protrusions by performing cell fitting on the binarized image data. Claim 2
The cell image analyzer according to item 1. 5. The analysis processing means compares the average value of the entire image or the average value of the DC components of the frequency components on the low frequency side with the neighborhood average value of the target block so that the target block is a cell part. The cell image analyzer according to claim 4, wherein the threshold value of the binarized image data is made variable by determining whether it is a medium or a medium. 6. A binarization processing step of binarizing image data of a cell image obtained by imaging cells of animals and plants to generate binarized image data, and frequency-converting the image data to generate frequency components. A frequency conversion step of band-dividing the high frequency side frequency component and a low frequency side frequency component containing a direct current component, and a cell block extraction step of creating intermediate data for feature extraction using at least the high frequency side frequency component And an analysis processing step of comparing the intermediate data and the binarized image data to extract a cell portion and a protrusion portion of an animal or plant. 7. In the cell block extracting step, one or more frequency components of the two frequency components are used to divide a frequency region of the frequency components to be used into a plurality of blocks, and the plurality of blocks are divided from the plurality of blocks. The cell image analysis method according to claim 6, wherein a cell block is extracted. 8. The cell image analyzing method according to claim 7, wherein the cell block extracting means removes the protrusion block from the cell block by utilizing the directional strength of the blocks and the arrangement pattern. 9. In the analysis processing step, when the cell block and the binarized image data are overlapped, if the binarized image data is present around the cell block, it is determined whether the cell block is a cell portion. It is determined that the medium is a medium when there is no binarized image data, and the cell portion is detected from cells having protrusions by performing cell fitting on the binarized image data. The cell image analysis method according to claim 7. 10. The analysis processing means compares the average value of the entire image or the average value of the DC components of the frequency components on the low frequency side with the neighborhood average value of the target block to determine that the target block is a cell part. 10. The cell image analysis method according to claim 9, wherein the threshold value of the binarized image data is made variable by determining whether it is a medium or a medium. 11. A recording medium on which a program for executing each step of the cell image analysis method according to any one of claims 6 to 10 is recorded.