JPH0378886A - Applierd digital analysis apparatus and method for multi-purpose image data - Google Patents

Abstract

PURPOSE: To attain the objective and exact measurement and analysis of applied images or samples by converting the images of prescribed samples into a digital matrix indicating a second dimensional plane, and into statistical divisions indicating the constitutes of the prescribed samples. CONSTITUTION: This device is provided with a means 10 which captures the images of the prescribed samples, means 16 which converts the images into digital matrix indicating a second dimensional plane, means 20 which improves the digital matrix, means 20 which processes the improved digital matrix, means 20 which converts the second dimensional digitized data into statistical divisions indicating the constitution of the prescribed samples, and means 26 which image- displays outputs including the statistical divisions. Thus, the particle structure of an organic substance can be analyzed without being affected by the subject of an operator, and without selecting an object to be measured and a measuring place.

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial Application Field] The present invention relates to digital image processing technology, and particularly to a quantitative analysis method and apparatus for a sample piece using digital image processing. In particular, the present invention applies image enhancement techniques, continuity analysis, and morphological analysis to an analysis target or pattern in an input image captured using an imaging device such as a television, camera, or video camera. This invention relates to an algorithm-based method and apparatus for analyzing samples related to palynology, sedimentology, etc.

[Prior Art] Conventionally, the particle structure of organic substances has been expressed in terms of size and shape by directly measuring and quantifying each particle visually observed through an observation device such as a microscope, or by comparing it with a visual chart. Ta.

Measuring pore diameters and analyzing inorganic substances in rocks has typically been carried out using mechanical devices that measure the number of components per unit area, or by measuring projected images using a manual area meter.

The results of these measurement methods depend on the subjectivity of the measurer, and are inefficient.

An early technique for quantitative characterization of the constituent materials of deposits was developed by M, A, 1 orete.
Papers by Palinology and Palinofacies
Of the Avatar in Venezuela (“P
ALYNOLOGY AND PALYNOFACIE
SOF T)IE UPPERTERTIALLY IN
VENEZUELA-). This paper discusses a method for extracting geologically useful information by processing raw data obtained by observing groups of organic particles obtained by the QUANTINET device.

This patent literature is characterized by several patents in particle analysis.

U.S. Pat. No. 4,229,797 discloses means and apparatus for automatically measuring particle composition and color by means of an electronic device by means of image analysis. The method is to smooth the initial image, that is, subtract the locally averaged secondary image by electrical processing, and use the result containing local extreme values as a kind of parameter. By using images analyzed up to the next and third orders as parameters,
Tables and standard patterns of composition in pre-measured and classified medical samples can be compared.

British Patent No. 2,029,570-A describes the structure and method of an apparatus for analyzing images and automatically classifying the area occupied by a target substance. In this device, an operator interactively inputs a threshold level of brightness in an image, thereby extracting a portion that the operator determines to be included in the target material from other materials within the field of view.

By automatically measuring and recording threshold levels and other parameters necessary for computer operation, the device allows the image of interest to be compared and classified with other already classified images. The application for this recognized patent specifically relates to the calculation of total porosity, the determination of shape and grain size distribution, the degree of connectivity, and the determination of the shape of sand grains in sediments and fractures. ) refers to a method using an image analysis device. The problem with this invention is that it requires very complex equipment.

The device cannot be changed and is inflexible.

Furthermore, the configuration of the classification device depends on the operator's judgment and has difficulty in reproducibility.

Further disadvantages include the high cost and relative lack of portability due to the complexity of the equipment, which is disadvantageous for operation in remote locations in geology. US Pat. No. 4,700,298 describes an image processing apparatus using a microscope for recognizing, measuring, and positioning living cells in a tissue cultured in a flask.

The content is limited to local positioning of cells within a two-dimensional field of view, and by measuring the shading of the image, cells can be recognized from among cell debris and other incomprehensible objects. No shape or shape-related elements are extracted.

In general, these systems are not suitable for quantitative density characterization of organisms or petrographic thin sections polished from sample rock fragments collected from oil fields.

Therefore, in the present invention, it is possible to determine the density m
We propose equipment and methods that can also be used for characterization.

Furthermore, the present invention proposes an apparatus and method that can also be used for samples containing morphological data using digital analysis techniques.

[Problems to be Solved by the Invention] In the prior art, there is a problem in that the measurement results are subject to the operator's subjectivity, and this was only achieved using a very expensive, complicated, and inflexible device. Therefore, an object of the present invention is to eliminate the subjectivity of the operator, and to
and to provide equipment and processes that allow measurement anywhere.

[Means and operations for solving the problem] In order to achieve the above object, a means for capturing an image of a predetermined sample, a means for converting the image into a digital matrix representing a two-dimensional plane, and a means for improving the digital matrix are provided. , means for processing the improved digital matrix, and means for converting the two-dimensional digitized data into statistical partitions representative of the composition of a given sample;
There is provided a digital analysis device for morphologically characterizing an image composition, characterized in that it includes means for displaying an output including said statistical classification as an image.

It is also possible to provide an apparatus in which the image capturing means comprises a video camera and the digitizing means comprises an electronic digitizer.

The means for improving the digitized image can also provide an apparatus comprising a microprocessor with predetermined software and a mouse device.

The microprocessor includes at least means for processing the enhanced image, and the microprocessor includes means for processing the enhanced image mathematically and logically to determine a morphological characterization of a given specimen. Apparatus can also be provided to perform deeper measurements of constructs.

An apparatus may also be provided, characterized in that said microprocessor is processed to carry out said mathematical and logical operations, and said operating means further include means for inputting manual input instructions to said microprocessor.

An apparatus may also be provided, characterized in that it includes a first video for displaying the captured and improved image on a screen.

It is also possible to provide an apparatus characterized in that it includes means for modifying each image by editing the image.

It is also possible to provide an apparatus characterized in that the correction means includes a mouse-type indicator for inputting deletion of constituents in the image.

It is also possible to provide an apparatus characterized in that the improvement/modification and conversion means includes a microprocessor and a manual interactive operation input means to the microprocessor.

It is also possible to provide a device characterized in that the captured image is at least an image of a microscope or one of a series of photographs.

Further, capturing an original image from at least one of a microscope or a photograph, electronically converting said original image into a digital matrix representing a two-dimensional plane, transmitting said digitized image to a video monitor, and transmitting said digitized image to a video monitor; and processing the two-dimensional data of the digitized image into statistical divisions,
There may be provided an image digital analysis process for apothetical characterization of an image composition, characterized in that the digitized image is analyzed and an output of the statistical partitioned image is produced.

It is also possible to provide a process characterized in that said image capture means are always able to view images from said microscope or video camera.

It is also possible to provide a process characterized in that the correction means includes means for improving the digitized image by erasing misidentified structures or adding input of structures that could not be captured.

The improvement method can also provide a process characterized in that the improvement method further includes correcting the human output gray level of the image using a hardware input/output gray level correspondence table.

It is also possible to provide a process characterized in that the modification method includes a method of limiting and selecting a comparison according to the gray level and manipulating the image according to the selected gray level to extract the gray level.

A process may also be provided, characterized in that the processing method comprises further digitizing by delimiting the composition of the image by at least mathematical and logical operations.

The analysis method may include determining morphological parameters by determining at least the periphery of a certain area of each image constituent, relative elongation rate, unrestrictedness, and inclination from the density of the target substance. Can be provided.

It is also possible to provide a process characterized in that the processing method further includes determining the boundary of the image composition by an 8-connection method.

What is the 8-connection method? In a two-dimensional pixel matrix, if two pixels are connected by edges and vertices, there are eight ways to connect them.

A process may also be provided comprising a display monitor and a memory for storing a digitized image and a method for displaying the desired improved digitized image on said monitor.

The analysis method converts the processed data into the statistical classification, classifies the size of the image components by diameter classification, and statistically calculates the relative elongation rate of each component extracted as a sample. Characterization by classification characterizes the irregularity of each of the structures by comparing its circumference with the circumference of a circle with the same area as the structure, and the inclination of the structure is determined by the semicircular projection of the structure. It is also possible to provide a process characterized by determining.

It is also possible to provide a process characterized in that it includes a process of accumulating the total area of similar parts of the image constituent parts and calculating the area and percentage of the remaining parts.

It is also possible to provide a process characterized in that it always processes each digitized image and also stores the processed data of each digitized image.

A process can also be provided, characterized in that it uses directly said image output by statistical partitioning on the morphological features of the sample image.

A process can also be provided, characterized in that it uses directly said image output by means of statistical classification in morphological characteristics of organisms.

A process can also be provided, characterized in that it directly uses said image output with statistical classification in some morphological features in rock taxonomy.

A process may also be provided, characterized in that the image output by statistical partitioning is used directly in the interpretation of quantitative and morphological characterizations other than the extraction part of the sample image.

A process characterized in that it includes a method for automatically classifying rocks according to mineral ratios can also be provided.

[Example] The present invention relates to an apparatus and method for extracting morphological features in the composition of an image using digital analysis of the image.

This device and method quantitatively analyze features in two-dimensional grayscale images in statistical measurements of morphological characteristics and the degree of particle dispersion in images.

The apparatus and techniques also broadly include techniques for digitizing and recording images obtained by television cameras. The image analysis device in this advertisement has a resolution of 512X4.
It has 80 pixels and 128 gray levels, and the device also includes a display that displays the gray level of the captured image at high resolution.

The device further includes a method of using digitized images;
This includes a method for modifying the image, a method for statistically processing the two-dimensional original image so that the structure of the sample can be characterized, and a method for outputting a display image containing the results of the statistical processing. .

The output includes statistical curves and elements to characterize the sample image. Image processing includes techniques for removing and editing unnecessary parts of images.

The apparatus in this application required for image processing and processing of original images is a general-purpose microprocessor that is fully equipped with a process duram and an operator interface necessary for the desired processing in advance.

The present invention also includes a processing method that is obtained by manipulating the level of shading of a portion of the currently captured image that is desired to be emphasized.

The user can select the desired portion of the image.

Quantification can be performed by binarizing a defined image before performing spatial processing on the data. An algorithm based on the 8-connection method allows contour extraction to be applied to objects in the required part of the image, and the ij element can be calculated for each object.

The present invention provides an apparatus and method for quantitatively analyzing the constituent substances of a two-dimensional grayscale image, and further measuring the statistical dispersion and morphological characteristics of target particles in the captured image by comparing with previously generated grayscale information. It is related to.

Generally speaking, the method and integrated device in the present invention are shown in Figure 1.
．． Shown in 2. This includes a series of image capture units by a television camera, IO in the figure. The images are a series of specimen photographs 12 or through the lens of a microscope 14 . A monochrome video camera such as the R5-170 is used as the camera IO because it has a higher resolution than a color camera. By using such a camera, it is possible to obtain digital image information with 512×480 pixels and 128 gradations.

Although not shown in the figure, the camera IO may use a lens system or an aperture to adjust the focus and amount of light.

Optimization of lens systems and apertures is a matter of art and is beyond the scope of discussion.

The electronic digitizer 16, that is, the captured image 512
The device converts the video signal obtained by the television camera 10 into information that can be captured by the device.

Digitizer 16 is a device well known in the circuit art for converting partial relative brightness values captured by a camera into a digital matrix represented in two-dimensional virtual coordinates.

The digitized image is transferred into memory 18 in microprocessor 20. A first high resolution video display monitor 22 displays images stored in a digitizer obtained from a TV camera 10, such as a CRT connected to a digitizer 16, or images stored in a microprocessor 20. indicate.

The microprocessor 20 may be a card or module equipped with programs for various controls and functions, or any suitable computer used in the industry.

For example, a microprocessor has 512KB RAM
, a desktop computer such as an IBM-PC/AT equipped with a 20 MB hard disk and a color monitor 26 is used. The microprocessor 20 may be equipped with a mouse 24 or a light painter interface to simplify editing of processed images. Microprocessor 20 can number the generated images.

It is also possible to perform arithmetic or logical operations such as transfer, comparison, addition, subtraction, correlation, and multiplication on the converted darkness and light brightness of each of the captured images.

Before performing quantitative analysis of the image, the device must be configured for determining the correspondence between the actual information and the measured pixels. First of all, a reference with high brightness in length and horizontal and vertical directions must be constructed.

Measurements may be made by changing the dimensions until the measurement field completely covers the reference. Also,
Preferably, the arrangement is made at each magnification of the microscope 14.

The above configuration method may be prepared in the microprocessor 20 as an appropriate program. Once constructed, a linear correlation will depend on the length between two actual points and the length between two corresponding points in the virtual coordinates of the actual image and in the digitized image. Information regarding these items can be taken into the microprocessor 20 and kept as a memory, or used for starting up the apparatus or for generalization in mutual quantitative analysis of images.

The configuration of the device facilitates computer analysis and allows for anticipated changes in camera positioning, size, shape, field of view, and the use of different camera and lens systems.

After initial setup and configuration, the target sample can be tested. For example, a sample is placed under the microscope 14. Focus the microscope or camera until a clear image of the sample is obtained on the high resolution monitor 22.

Analog images of the sample are continuously generated with the sample in view with the camera. The frequency of image data generation by a video camera is generally 30 times per second. The image captured by the camera lO is digitized by a digitizer 16 and sent to the microprocessor memory 18 and video monitor 22.
be taken in. The operator selects several useful image processing procedures for quantitative particle analysis and differentiation in particle images. The processing methods that can be selected are: (1) Adjustment of brightness and contrast in a pale image.

(2) Setting the threshold.

(3) Image selection.

(4) Editing images manipulated by threshold values.

(5) Quantitative analysis processing.

Menus extracted from the program interface for implementing the various steps of the above processing technique are shown below.

ADIE Main Menu Environment Menu A, Device Initialization B, Device Configuration C0 Work Area 1 ■ Processing Menu A, New Job B, File Reading C0 Menu for Consideration of Pollen, etc. l1 Menu for Image Editing Threshold Output Shading Scale Adjustment Particle removal Particle separation Capturing a new image 2, Image processing menu Menu for image editing Threshold output Concentration scale Particle removal Particle separation Capturing a new image Image processing File reading File writing New processing Graphic menu Creating a histogram of the particle distribution area Relative elongation rate Relative unrestrictedness 3, Measurement menu for consideration of pollen, etc. Maceral selection Threshold value Output Concentration scale adjustment Particle removal Particle separation Capturing new images Processing images Read file Reading file writing New processing Graphic menu Histogram of particle distribution area Relative growth rate Relative unrestrictedness, Image processing menu 1 for sediment, Menu for image editing Threshold Output Gray scale adjustment Particle removal Menu for image editing Particle separation Capturing a new image 2. Characterization Menu Hole Processing Menu Image Processing Image Editing Menu Threshold Output Concentration Scale Particle Removal Particle Separation Capture New Image Image Processing Read File Write New Process Graphic Menu Histogram Particle Distribution Area Relative Elongation Relative Unregulated Unit Particle Processing Menu Image Editing Menu Threshold Output Concentration Scale Adjustment Particle Removal Particle Separation New Image Capture File Read File Write New Processing Graphical Menu Histogram of Particle Distribution Area Relative Elongation rate Relative unregulated Unit 3, Menu for sediment measurement Menu for mineral measurement Menu for mineral selection Image editing menu Threshold output Gray scale adjustment Particle removal Particle separation Capture new image Process image Read file Write new file Selection of Processing Minerals Selection of the processing procedure is made by the operator, and the microprocessor 20 performs the necessary arithmetic and logical operations to carry out the selected processing procedure, e.g. When "point-to-point conversion" is performed using input from a digitizer, the center gray level and the gray level width of the band to be captured or the band to be discarded are selected.

Moreover, when an operator edits, necessary processing is executed according to the operator. Operations are entered by an operator and results are displayed directly on monitor 22 or 26. The input device is
This can be done with an existing mouse type pointer 24 or a light pen 28, which allows useful procedures to be manually set.

In this mode, the keyboard 30 is used to input instructions for performing certain operations.

Adjustment of the brightness level and contrast in the image data captured by the digitizer 16 is performed by "point-to-point conversion." If a linear transformation with a negative slope is performed, an image with innocence can be obtained.

Using this technique, it is possible to quantitatively measure spaces and porosity using the same technique as for particle measurements. One of the features of the present invention is that the operator can operate continuously while checking the results between processes.

To facilitate the extraction and characterization of objects in images,
The selection of the threshold level that requires setting a certain threshold level is manually set by the operator, and since the target object can be sorted out from the digitally quizzed image sample by the threshold level, the influence of the quiz is not included as much as possible. In addition, a value that more clearly distinguishes the object is required. Quantitative analysis can only be performed on images manipulated by threshold values.

If necessary, the threshold level can be updated by the operator. Furthermore, the input and output gray levels can be corrected by creating a correspondence table of input and output gray levels.

Image editing allows unnecessary trajectories, spots, and objects to be excluded from the field of view of the digitized image, allowing them to be included or excluded from the quantitative analysis results. Editing has range and line modes, which are used for rough editing and fine editing, respectively.

Editing, like erasure, can characterize an image. As a matter of course, image editing is performed using the mouse 24 and the light pen 28.

In some cases, it may be advantageous to use an optical filter 23 and its support 21.

It is best to use a filter only when the density level of the target substance is very specific.

For example, the operator uses the mouse pointer 24 to set the center of the required gray level. Using the operation keys of the microprocessor 20, the operator selects whether or not to operate the filter, the characteristics of the filter, and the band to be taken in or the band to be discarded.

After all corrections on the image sample are completed, the operator proceeds to a processing routine to obtain morphological characteristics based on quantitative analysis, statistical characterization, etc. of the image sample. Image processing is performed within the required image area. During processing,
Within the required area, each line is observed in detail and positioned relative to the extracted group based on the threshold level set by the operator.

This group has a correlation with the measurement sample, and its morphological characteristics are
It is later extracted by a digital processing algorithm. This measurement sample can be called a "target substance."

Once the object is located, its contour pixels can be traced using a technique known as the 8-connection Turdle algorithm.

What is the 8-connection method? In a two-dimensional pixel matrix, if two pixels are connected by edges and vertices, there are eight ways to connect them.

FIG. 3 shows the processing procedure of the algorithm. First stage,
The target substance is input horizontally.

It is classified as type 3 as shown in Table I. In the second stage, the target substance is left behind and exits in the vertical direction to be classified as type 5.

This is taken over by input from the horizontal direction like type 4. Steps 1 to 8 zigzag the inner and outer surfaces of the target material. This is an algorithm that follows the contour and traces the boundaries of the target material.

The boundaries of the target material are "marked" by arbitrarily set binary codes, rather than by the range of gray levels of the digitizer image. By marking the boundaries like this,
Image operations can be performed without reprocessing already processed areas.

The algorithm that traces this boundary is a “weak”
It is designed to skip connections. In the 8-connection method, a case where two pixels are diagonally connected can be called a weak connection. In a more general environment, in an erosion operation, the circumference of particles in an image gradually decreases, sometimes breaking strong connections, i.e., 8-connections to one or more pixels. .

Each pixel is centered in the digitized coordinates or
Rather, it is recognized by virtual coordinates relative to the reference point of the corner. When implemented, the digitized image is represented by a rectangular array in two dimensions.

That is, each pixel consists of 4 pixels, two on the horizontal side and two on the vertical side.
It has two sides. Each object is represented by an inner region and a boundary pixel. Boundary pixels are characterized by having inner and outer faces.

The inner surface corresponds to two pixels belonging to the same object connected by the same surface. The outer surface corresponds to the connection surface between the boundary pixel of the target substance and its surroundings.

The inner and outer surfaces are extracted using an algorithm that traces the outline of the boundary of the target material. The outer surface is used for cumulative processing of the target substance as a region element.

As mentioned above, the intersection of the outer planes is characterized by position and perception of movement, if movement occurs. In addition,
Intersection of the outer surfaces means that the target substance must be left in its surroundings or that the target substance must be reduced to its surroundings. Similarly, there are eight ways of intersecting the outer surfaces, each of which is shown in the table.

Table: Intersection style type of outer surface Direction Sense Generates virtual pixels that separate the correlation field with the target substance. Based on that information and knowledge of the intersecting pattern of the outer surfaces, the following quantities are calculated by accumulating:

SUMX: Sum of relative X coordinates SUMY: Sum of relative Y coordinates SUMXY: Sum of relative XY coordinate correlations (relative X coordinate x
Relative Y coordinate) 30MX2: Square root of relative X coordinate SUMY2: Square root of relative Y coordinate The operator can perform the above calculation for all pixels containing the target substance.

Let Npxs be the number of pixels included in the target substance, and define the element M [i, j] of the array variable regarding the target substance as follows.

M[0,1 piece: SUMY/Npxs
(1) M[1,O
] : SUMX/Npxs
(2) M[0,2]: SUMY2-(SUMY)”
/Npxs (3)M[2,0]:
30MX2-(SUMX)”/Npxs
(4) M[1,1]: SUMY-(SUMX
SUMY)/Npxs (5) The boundary length of the contour traced by this method can be found by simple cumulative calculation. Other morphological features regarding the target substance are also derived from the elements of the array variables described above as follows.

Slope = 0.5 Arctangent (
2xM[1,t]/(M[2,0]-M[0,2]))
' (6) Elongation rate (elongat 1on
) = (Mp[2,0/Mp[0,2])""
(7) However, Mp[2,0]=M[0,2]x(sine(slop
e) door + M[2,0]x(cosine(slope
))+2x(sine(slope))x(cosin
e((slope))Mp[0,2] = M[2,0
]x(sine(slope))+M[0,2]x
(cosine(sloipe))” 12x(sloipe)
(slope)) x (cosine (slope)) area (area) = Npxs (
8) Irregularity = perimeter/
Perimeter converted into a circle of equal area (9) However, the circumference converted into a circle of equal area is the circumference of a circle having the same area as the area of the target substance.

Image components are classified by the distribution of phi diameters.

phi is a number expressed in log'' (particle diameter).

The morphological elements of the object in the field of view of the digitized image can be handled by a file system. The results of analyzing disparate images can be combined and treated, thereby allowing meaningful statistical results to be drawn from a sample group. Specifically, the statistical analysis results can be graphed and displayed as distribution, accumulation, etc.

In addition, new statistical element analysis methods can be added, and calculation results can be displayed on the display. FIGS. 4 to 7 are diagrams showing some results of statistical processing. FIG. 7 shows the directionality of the constituent elements on a semicircular graph.

Similar to the output of images, charts and tables can be output on the monitor 26 or on a printer (not shown in the figure). The resulting graphs of statistical distributions generated by this system can be used directly for the following purposes:

(1) Morphological characterization of living organisms in the sample, (2) Morphological characterization of mineral samples in mineral taxonomic sample thin sections, (3) Spaces (holes) in the sample image by image brightness inversion. morphological characterization and quantification; (4) classification of minerals in samples using the same method as mineral classification and automatic classification of rocks based on their percentages;

Furthermore, the area calculation process can calculate the total area of the same object group, and also calculate the percentage of the area of the object group relative to the remaining area. These techniques are described in the article by Lorente mentioned above.

It is understood that the invention encompasses more than just the apparatus and description herein. The device diagram shown here is merely a form that is easiest to express the present invention, and its shape, size, constituent parts, and details of operation can be easily modified. . The present invention includes all modifications and variations of the device that fall within the scope of the appended claims.

[Effect] The apparatus and method of the present invention enable objective and accurate measurement and analysis of a given image or sample,
Furthermore, the range of measurement targets is expanded, and the measurement location can be flexibly handled.

[Brief explanation of drawings]

FIG. 1 and FIG. 2 are configuration diagrams of the device in the present invention,
FIG. 3 shows the processing procedure of a contour extraction method known as the Turdle algorithm. 4, 5, 6, and 7 are some of the results generated using the apparatus and techniques of the present invention. 10, Camera 12, Sample photo 14, Microscope 16° 18° 20° 22° 24° 26° 28° 30°, digitizer, memory, microprocessor, video display monitor, mouse, color Monitor, light pen, keyboard 51 RELATIVE IRREGLILARITYNLI
MBEROF PARTI (LES = +oo9RE
LATIVE IRREGLILARITYVALL
IE (VIo 5CALE) JFZG-41 J'7G-5

Claims (1)

Claims: (1) means for capturing an image of a given sample; means for converting the image into a digital matrix representing a two-dimensional plane; means for improving the digital matrix; and the improved digital matrix. an image configuration characterized by comprising: means for processing the two-dimensional digitized data into statistical divisions representing the composition of a predetermined sample; and means for displaying an output including the statistical divisions as an image. A digital analysis device that performs morphological characterization of objects. 2. The apparatus of claim 1, wherein said image capturing means comprises a video camera and said digitizing means comprises an electronic digitizer. 3. The apparatus of claim 2, wherein the means for improving the digitized image includes a microprocessor with predetermined software and a mouse device. (4) the microprocessor includes at least means for processing the enhanced image, and the microprocessor mathematically manipulates the enhanced image to determine a morphological characterization of a given sample; 4. Apparatus according to claim 3, for performing deep measurements such as image constructions for the purpose of image construction. (5) said microprocessor is programmed to perform said mathematical and theoretical operations, and said operating means include:
5. The apparatus of claim 4 further comprising means for inputting manual input instructions into the microprocessor. The apparatus of claim 2, further comprising: (6) a first video monitor for displaying the captured and improved image on a screen. 7. The apparatus of claim 1, further comprising means for modifying each image by editing said image. (8) Claim 7, wherein the modification means includes a mouse-type pointing device that inputs deletion of components in the image.
Apparatus described in section. 9. The apparatus of claim 8, wherein said improvement modification and conversion means includes a microprocessor and manual interactive operation input means to said microprocessor. 10. The apparatus of claim 1, wherein the captured image is at least an image of a microscope or one of a series of photographs. (11) Capture an original image from at least one of a microscope or a photograph, electronically convert the original image into a digital matrix representing a two-dimensional plane, transfer the digitized image to a video monitor, and transfer the digitized image to a video monitor; modifying the image and processing the two-dimensional data of the digitized image into statistical partitions;
An image digital analysis method for performing a morphological feature mapping of an image composition, characterized by interpreting the digitized image and generating an output of the image of the statistical division. 12. The method of claim 11, wherein the image capture technique includes constantly viewing the image of the microscope or video camera. (13) The correction method is a method of improving the digitized image by erasing misidentified components or adding input of components that could not be captured.
The method described in Section 2. (14) The method according to claim 13, wherein the improvement method further includes modifying the input and output gray levels of the image using a hardware input and output gray level correspondence table. (15) The correction method includes a method of limiting and selecting the comparison based on the gray level, and manipulating the image based on the selected gray level to extract the gray level. The method described in section. 16. The method of claim 11, wherein the processing technique includes further digitizing by delimiting the composition of the image by at least mathematical logical operations. (17) The interpretation method is characterized in that the interpretation method includes determining morphological parameters by determining at least the periphery of a certain area of each image constituent, relative elongation rate, irregularity, and inclination from the density of the target substance. 17. The method of claim 16. 18. The method of claim 16, wherein the processing method further includes determining the boundary of the image composition using an 8-connection method. 19. The method of claim 12, further comprising a display monitor and a memory for storing the digitized image and displaying the desired enhanced digitized image on the monitor. (20) The interpretation method converts the processed data into the statistical classification, classifies the size of the image components by diameter classification, and extracts the relative elongation rate of each component as a sample. The irregularity of each of the structures is characterized by comparing its circumference with the circumference of a circle with the same area as the structure, and the inclination of the structure is determined by the statistical classification of the structure's semicircle. Claim 1, characterized in that the determination is made by projection.
The method described in Section 1. (21) The method according to claim 11, further comprising a method of accumulating the total area of the similar parts of the image constituent parts and calculating a percentage of the total area with respect to the area of the remaining parts. (22) always process each digitized image, and
21. The method of claim 20, further comprising storing processed data for each digitized image. 23. A method according to claim 11, characterized in that the image output by statistical partitioning is used directly in the morphological features of the sample image. 24. A method according to claim 11, characterized in that the image output by statistical partitioning is used directly in the morphological characteristics of the organism. (25) The method according to claim 11, characterized in that the image output by statistical classification is directly used for some morphological features in rock taxonomy. (26) The method according to claim 11, characterized in that the output from the statistical classification is directly used in the interpretation of quantitative and morphological characterizations other than the extraction part of the sample image. (27) The method according to claim 21, which includes a method of automatically classifying rocks based on mineral ratios.