JP2730435B2 - Measurement method of circular particles - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a method for measuring the diameter of a circular particle.

[0002]

2. Description of the Related Art Particles of solid, liquid, or gas having a substantially spherical or circular cross section are projected on a two-dimensional screen to form an image, and the diameter of each particle is measured to obtain the particle size distribution. It is necessary to measure the diameter of Since the actual image of the particles is not a perfect circle, the area or the perimeter is determined from the image and the diameter is calculated as a circle according to the purpose, or the maximum scanning length of the particles is used as the particle size according to the purpose. Was. However, when the particles on the image exist independently of each other, or when they are in contact with each other, the particle size of each particle can be easily measured. However, it is difficult to measure the size of each particle in a state where the particles overlap each other, that is, in the case of a photographed image that is referred to as a coagulate particle or an aggregate particle.

FIG. 6 shows an image pickup screen of a set of circular particles.
Part a shows a state in which two particles are overlapped, part b shows a state in which four particles are overlapped, and part c shows a state in which many particles are in contact with little overlap.

FIG. 7 shows the state of FIG. 6 in which each particle is separated by a known method. When there is little overlap as in part c of FIG. 6, each particle is separated while maintaining a fairly circular shape. However, when the overlap between the two particles is large as in part a of FIG. 6, the separated boundary is a straight line, and the shape of the separated particles is a kamaboko shape. When four particles overlap as shown in part b of FIG. 6, the shape of the particles after separation is quite different from a circle.

[0005]

In FIG. 7, when measuring the diameter of a particle, conventionally, a substantially circularly separated particle is extracted and its diameter is measured. As a result, the overlapping particles are excluded from the object to be measured, so that the number of particles that can be measured at one time decreases, and it is necessary to increase the number of visual fields to be observed. Therefore, when measuring the particle diameter and creating a histogram or the like, it takes a lot of time and effort, and it takes time to measure. There is a problem that particle size distribution data cannot be obtained in the particle distribution state.

SUMMARY OF THE INVENTION The present invention has been made in view of the above problems, and has as its object to provide a method for measuring circular particles capable of obtaining the particle size of overlapping particles.

[0007]

In order to achieve the above object, a set of circular particles is imaged, and each particle is subjected to binary image processing.
Child performs image TDG until one pixel, then the
A process is performed in which a single pixel is centered and circularly thickened to its original size. At this time, a predetermined pixel is vacant even if the boundary between the particles is minimum, and between any two points on the circumference of each particle after this thickening process. Is the diameter of each particle.

In the thickening process up to the original size, a logical product of an image slightly thicker than the original size and an image of the original particle is obtained.

[0009]

The image of the overlapping circular particles is reduced by binary image processing until each particle becomes one pixel.
The particles are thickened in a circle around the one pixel to the original size. The original size is circular because it is thicker than one pixel
Can be restored. Also, in the process of thickening, the overlapping particles are separated by making the borders of each other thicker so that the minimum number of pixels predetermined is empty. After the fattening process, the maximum length between any two points on the circumference of each particle is defined as the diameter of the particle. This is because the measurement target is a circular particle, and because it overlaps, even if the circular thickening process cannot be performed, the maximum length between any two points on the circumference of the restored particle is the same as the original particle. Based on the finding that it is almost equal to the diameter. As a result, the diameter of particles that are not restored to a circular shape can also be measured, and it is not necessary to remove overlapping particles in one field of view from the measurement target as in the related art.

When the fattening restoration process is performed, an image obtained by restoring the original image by one pixel to several pixels larger than the original size and the original captured image are logically ANDed, so that except for the boundary between adjacent particles, The original image can be accurately restored.

[0011]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram of an image processing apparatus for realizing an embodiment of the present invention. An A / D converter 2 for converting input data from the television camera 1 from analog to digital; an input buffer 3 for storing the converted data;
CPU 5 for overall control, image processing processor 6 for processing image data, program memory 7 for storing a program for determining the operation of CPU 5, image processor 6
Are a binary image memory 8 for storing the data subjected to the binarization processing and the gradation processing, a gradation image memory 9, an output buffer 10 for temporarily storing the output data, and a D / A for converting the digital data of the output data into analog data. converter
11, composed of a CRT 12 for displaying output data.

FIG. 2 is an overall flowchart of this embodiment. First, a target image is input from the television camera 1 (step S
1). In the case of powder, fine particles in a liquid, or small foreign matter in a solid, the television camera 1 is attached to a microscope to take an image. Shading correction is performed on the input image (step S2). Here, the shading correction is a process of correcting a non-uniformity due to illumination of an imaging target or an imaging device, or a gradual change in an original image signal level caused by a change in a background. Next, binarization processing is performed (step S3), and the image data shown in FIG. 6 is obtained. The binarized data is separated into particles by a method described later (step S4).
The result of the separation is as shown in FIG. 7, and those having many overlaps do not become circular.

The frames shown in FIGS. 6 and 7 show the boundaries of one field of view. In FIGS. 6 and 7, the particles are all contained within this frame. However, if there are any particles that do not fit all of the particles in this frame, they are excluded from the measurement target. This is called edge removal (step S5). Labeling processing is performed on the separated images as shown in FIG. 7 to extract particles (step S
6). Further, fine particles and dust less than the predetermined area are excluded from the measurement target (step S7). The maximum diameter of each particle, that is, the maximum length between any two points on the circumference of each particle is measured from the image data shown in FIG. 7 obtained as described above, and this is set as the diameter of the particle (step S8). ). Particle size distribution data is created from the data thus obtained (step S9).

FIG. 3 is a flowchart of the particle separation process, and FIG.
FIG. 3 is a diagram showing a situation where particles are separated according to this flow chart. In order to perform this separation processing, image reduction and enlargement processing is performed, but this processing itself is a known technique. For example, “Image processing subroutine package, SPIDER USER'S MANU
AL, supervised by the Industrial Technology Institute, published by Kyodo System Development Co., Ltd. "
And so on.

When performing reduction and enlargement processing, logical filter processing is performed. In this process, the 3 × 3 pixel processing area of the binary image is moved from the upper left to the right one pixel at a time to the right end, and when it reaches the right end, it is similarly moved from the left end one line below to the right end one pixel to the right. Scan the screen. Each time one pixel moves, it is determined whether the center pixel is set to 1 or 0 based on a 3 × 3 pixel pattern (256 types), and the center pixel of the processing area of 3 × 3 pixels is replaced. .

FIGS. 5A and 5B show a reduction process and an enlargement process. FIG. 5A shows a reduction process, and shows a case where the image is reduced to two points indicated by diagonal lines (this point is referred to as a center). (B) shows a state where enlargement processing is performed based on these two centers. In order not to connect the two figures, a minimum of one pixel gap is provided for eight neighborhoods (eight surrounding pixels including a diagonal), and the figure is restored to the original size. In this drawing, eight neighbors are shown, but a clearance may be provided for four neighbors (four vertical and horizontal pixels).

In FIG. 3, a binary image is input (step S1). FIG. 4A shows this binary image. The binary image is reduced using a logical filter for reduction (step S2), and reduced to one pixel at the center (step S3). The number of reductions N when the center pixel becomes 1 is 1
Is obtained (steps S4 and S5). The situation at this time is shown in FIG. Next, enlargement processing is performed using an enlargement logical filter based on the two centers (step S6). The enlargement process is performed once more than the number of times of the reduction process, that is, the image is enlarged to a size larger by one pixel than the original figure (step S7). This situation is shown in FIG. Next, the logical product of the original image shown in FIG. 4A and the enlarged image shown in FIG. 4C is calculated (step S8). As a result, FIG.
(D), an image separated by the size of the original image is obtained (step S9). By taking the logical product in this manner, the shape of the original image is reproduced with high fidelity.

[0018]

As is clear from the above description, in the case of the present invention, in the case of a circular particle, the overlapping particle is reduced to one pixel.
Conventionally, the shape after separation is not circular by setting the maximum length of any two points on the circumference of each particle which has been enlarged and separated and processed to be a circle with the original size as the diameter of the particle. Therefore, the discarded data can be used as valid data. This saves labor and time of measurement,
Further, since data is less often discarded, data bias can be prevented.

[Brief description of the drawings]

FIG. 1 is a configuration diagram of hardware for realizing an embodiment of the present invention.

FIG. 2 is an operation flowchart of the embodiment.

FIG. 3 is an operation flowchart of a particle separation process.

FIG. 4 is a diagram illustrating a particle separation process.

FIG. 5 is a diagram showing a reduction process and an enlargement process.

FIG. 6 is a diagram showing an original image of a circular particle.

FIG. 7 is a view showing a separated image of the original image shown in FIG. 6;

Claims (2)

(57) [Claims] An image of a set of circular particles is imaged, image degeneration is performed by binary image processing until each particle becomes one pixel, and then a circle is formed around this one pixel to its original size. The thickening process is performed. At this time, a predetermined pixel is vacated even if the boundary between the particles is minimum, and the maximum length between any two points on the circumference of each particle after the thickening process is defined as the diameter of each particle. A method for measuring circular particles. 2. The method according to claim 1, wherein the thickening process to the original size is performed by taking the logical product of the image slightly thicker than the original size and the image of the original particle. Circular particle measurement method.