JPH01238652A - Object recognition method for radiographic image - Google Patents

Abstract

PURPOSE:To improve detection accuracy of the limits of regions and to improve image quality of reproduced images by dividing a radiographic image into small regions so as to permit the dispersion value to be the same or less than a specified value, unifying the small regions by evaluating the similarity and obtaining characteristic values. CONSTITUTION:The radiographic image 21 composed of an object part 22, a direct radiation part 23 and a scattered line part 24 is divided into small regions so as to permit, for example, the area to be (l)X(m). Then, the dispersion value of the image data of the each small region is compared with a specified value, and the image 21 is divided again over the all regions until the dispersion value become equal to or less than the specified value. The similarity of the adjacent regions is evaluated by the corresponding image data, and when there is similarity, the small regions are unified. Such operation is repeated over the all regions of the image 21 and the small regions are unified into plural regions which are not similar to each other. Then, more than one characteristic values of the image data are obtained by the region, and the object part 22 is recognized by the characteristic value.

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention is directed to reading a radiation image in which a subject is partially recorded and obtaining a large amount of image data carrying radiation image information. The present invention relates to a radiation image subject recognition method for recognizing an area occupied by a subject in a radiation image.

(Prior Art) It is practiced in various fields to read a recorded radiation image to obtain image data, perform appropriate image processing on this image data, and then reproduce and record the image. For example, an X-ray image is recorded using an X-ray film with a low gamma value designed to be compatible with later image processing, and the X-ray image is read from the film on which it is recorded and converted into an electrical signal. After converting and performing image processing on this electrical signal (image data), contrast and sharpness can be improved by reproducing it as a visible image in photocopies, etc.

Efforts have been made to obtain reproduced images with good image quality performance such as graininess (see Japanese Patent Publication No. 81-5193).

In addition, the applicant has proposed radiation (X-rays, α-rays).

When irradiated with β rays, γ rays, electron beams, ultraviolet rays, etc., a part of this radiation energy is accumulated, and then when irradiated with excitation light such as visible light, stimulable fluorescence exhibits stimulated luminescence depending on the accumulated energy. Radiographic image information of a subject such as a human body is partially recorded on a sheet of stimulable phosphor using a stimulable phosphor, and this stimulable phosphor sheet is scanned with excitation light such as a laser beam. The resulting stimulated luminescent light is read photoelectrically to obtain image data, and based on this image data, a radiation image of the subject can be recorded on a recording material such as a photographic light-sensitive material, a CRT, etc. A radiation image recording and reproducing system that outputs visual images has already been proposed (Japanese Unexamined Patent Publication No. 5
No. 5-12429, No. 5B-11395.

No. 55-11340, No. 5 [1-1041345,
No. 55-116340 etc. ).

This system has the practical advantage of recording images over a much wider range of radiation exposure compared to conventional radiographic systems using silver halide photography. In other words, in a stimulable phosphor, it is recognized that the amount of emitted light that is stimulated to emit light due to excitation after accumulation is proportional to the amount of radiation exposure over an extremely wide range.
Therefore, even if the amount of radiation exposure varies considerably due to various imaging conditions, the amount of stimulated luminescence emitted from the stimulable phosphor sheet can be read by the photoelectric conversion means by setting the reading gain to an appropriate value. By converting the radiation image into an electrical signal (image data) and using this image data to output the radiation image as a visible image to a recording material such as a photosensitive material or a display device such as a CRT, it is not affected by fluctuations in radiation exposure. Radiographic images can be obtained.

In the above system, the stimulable phosphor sheet is scanned in advance with a low-level light beam before obtaining image data by reading it under optimal reading conditions depending on the dose of radiation irradiated on the stimulable phosphor sheet. Pre-reading is performed to read the outline of the radiation image recorded on this sheet, the pre-read image data obtained by this pre-reading is analyzed, and then the sheet is irradiated with a strong light beam and scanned to find the optimal image for this radiation image. There is a system that is configured to perform actual reading in which image data is obtained by reading under reading conditions.

Here, reading conditions are a general term for various conditions that affect the relationship between the amount of stimulated luminescence light and the output of the reading device during reading, such as reading gain, scale factor, or , the power of excitation light during reading, etc.

In addition, regardless of whether the system performs this prefetching or the system that does not, the obtained image data (including prefetched image data) is analyzed to determine the optimal image processing conditions when performing image processing on the image data. Some systems let you decide. The method for determining the optimal image processing conditions based on this image data is not limited to systems using stimulable phosphor sheets, but for example, image data is obtained from a radiation image recorded on a recording sheet such as a conventional X-ray film. It is also applied to the system.

By analyzing the image data (including pre-read image data) and determining optimal reading conditions and image processing conditions, it becomes possible to reproduce and output a radiation image that is highly suitable for observation.

(Problem to be Solved by the Invention) By the way, in one radiographic image corresponding to read image data (including pre-read image data), a subject that needs to be read and image processed in actual reading may be photographed. area around the subject (direct radiation area), where the radiation emitted during imaging was directly irradiated onto the sheet (direct radiation area); There are areas (scattered radiation areas) in the sheet that are shielded from radiation and are irradiated with only a slight amount of scattered radiation.

Of these, only the subject area is the area that needs to be read in the main reading in a system that performs pre-reading, or the area that requires image processing to improve the image quality depending on whether the system performs pre-reading or not. .

The other areas may be made, for example, solid black in the reproduced image so as not to be distracting when observing the subject. When determining reading conditions and image processing conditions, it may not be possible to obtain a reproduced image of the best quality unless the subject is recognized. For example, when calculating the average value of image data of a subject, if image data of other areas are also averaged, the values may be completely different.

However, until now, there was no known effective means for automatically recognizing the subject area, so the next best method was to recognize the area of the irradiation field, which is the combination of the subject area and the direct radiation area. Reading conditions were determined to obtain image data within the irradiation field, or image processing conditions were determined to perform image processing on the image data within the irradiation field. This is because the boundary between the irradiation field and the scattered radiation area changes quite rapidly on the image, so this change was automatically recognized and the irradiation field area could be determined (for example, 155968 (1986). By recognizing the irradiation field, it is possible to obtain fairly high-quality reproduced images for most images, but image quality can be further improved if the subject area can be recognized.

SUMMARY OF THE INVENTION In view of the above circumstances, it is an object of the present invention to provide a method for recognizing a subject in a radiographic image, which recognizes a region occupied by the subject in a radiographic image from image data.

(Means for Solving the Problems) The object recognition method of the present invention reads a radiation image in which the object is partially recorded to obtain a large number of image data carrying radiation image information, and then divides the radiation image. Find a variance value of image data for each of a plurality of small regions, further divide each of the small regions whose variance value is equal to or greater than a predetermined value into a plurality of small regions, and find a variance value for each of the divided small regions; Segmentation is repeated over the entire region of the radiographic image until the variance values of the many finally divided small regions are all below a predetermined value, and finally two adjacent small regions are divided among the many finally divided small regions. , the similarity of these two small areas is evaluated based on the image data corresponding to these two small areas, and if the two small areas are similar, the two small areas are integrated into one By further evaluating the similarity with adjacent small regions and repeating this similarity evaluation over the entire region of the radiographic image, the large number of finally divided small regions can be divided into multiple regions that are dissimilar to each other. The method is characterized in that one or more types of characteristic values of the image data are obtained for each of the plurality of regions, and the region corresponding to the subject of the radiation image is recognized based on the characteristic values.

Here, the image data includes pre-read image data obtained by pre-reading.

(Function) The method for recognizing a subject in a radiographic image of the present invention obtains a variance value of image data for each small region into which a radiographic image is divided, and further divides small regions where this variance value is equal to or greater than a predetermined value. , parts of the radiation image with little change are not finely divided and remain as large areas, and parts of the radiation image with large changes, such as near the boundary between the subject and the direct radiation part, are finely divided.

By repeating the operation of evaluating whether or not two adjacent small areas are similar among the large number of small areas divided in this way and merging them if they are similar, the large number of small areas described above can be divided. Because we integrated multiple regions that are dissimilar to each other, compared to a method that mechanically divides a radiographic image into smaller regions and integrates them, large regions that cannot be divided finely because there are fewer changes in the radiographic image remain as large regions. Since it is used to evaluate the similarity, calculation time is saved, and since areas near the boundary, etc., which change rapidly in the radiological image, are divided into smaller pieces, the accuracy of detecting the boundary is improved.

When the characteristic values of the above image data are obtained for each of the multiple regions integrated in this way, the characteristic values show almost unique values for each region, although there are some variations depending on the individual radiation images of the same type. . Therefore, in the present invention, one or more types of characteristic values of the above-mentioned image data are obtained for each of the plurality of regions integrated as described above, and each of the small regions having similar characteristic values is divided into a plurality of small region groups. This makes it possible to recognize which small region group corresponds to the subject portion from the characteristic values.

One or more types of characteristic values that are statistically most suitable for recognizing the object part are used by examining similar types of radiation images in advance. Characteristic values that can be used include mean value, 3 variance values, and median value.

There are mode values, maximum values, minimum values, etc.

By recognizing the subject portion in this manner, reading conditions and image processing conditions can be determined so as to obtain the best reproduced image.

(Embodiments) Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings.

First, referring to FIG. 5, the entire apparatus using the present invention will be described.

FIG. 5 is a perspective view showing an example of a radiation image information reading device using the subject recognition method of the present invention.

This device uses a stimulable phosphor (stimulable phosphor) that, as mentioned above, accumulates a portion of this radiation energy when it is irradiated with radiation, and then exhibits stimulated luminescence in response to the accumulated energy when it is irradiated with excitation light such as visible light. This is a device that uses a fluorescent phosphor.

In a photographing device (not shown), a subject such as a human body is irradiated with radiation and photographed, and a radiation image of the subject is stored and recorded on a stimulable phosphor sheet.

The photographed stimulable phosphor sheet 1 is set at a predetermined position in the radiation image reading apparatus shown in FIG.

The stimulable phosphor sheet 1 set in this manner is conveyed (sub-scanning) in the direction of arrow Y by a sheet conveying means 3 such as an endless belt driven by a motor 2.

On the other hand, the excitation light 5 emitted from the laser light source 4 is transmitted to the motor 1.
A rotating polygon mirror 6 that is driven by 3 and rotates at high speed in the direction of the arrow.
After passing through a focusing lens 7 such as an fθ lens, the optical path is changed by a mirror 8 and the sheet 1
, and main scanning is performed in the direction of arrow X, which is substantially perpendicular to the sub-scanning direction (direction of arrow Y). Stimulated luminescent light is emitted from the portion of the sheet 1 irradiated with the excitation light 5 in an amount corresponding to the radiation image information stored and recorded, and this stimulated luminescent light 9 is collected by a condenser 10. and is photoelectrically detected by a photomultiplier (photomultiplier tube) 11 as a photodetector. The light collector 10 is made by molding a light-guiding material such as an acrylic plate, and has a linear incident end surface 10a extending along the main scanning line on the stimulable phosphor sheet 1. The light receiving surface of the photomultiplier 11 is coupled to the output end surface 10b formed in an annular shape. The stimulated luminescent light 9 that has entered the light collector IO through the incident end face 10a travels through the interior of the light collector 10 through repeated total reflection, exits from the output end face 10b, and is received by the photomultiplier 11, The light amount of the stimulated luminescent light 9 carrying the radiation image information is determined by the photomultiplier 1.
Detected by 1.

The analog data S1 output from the photomultiplier 11 is amplified by the amplifier 16, and then sent to the A/D converter 1.
7, it is digitized with a predetermined recording scale factor. The digital image data S2 thus obtained is input to the memory 18 and stored. memory 18
The image data S2 stored in is then read out by the calculation unit 19, and as described later, from the image data S2,
The subject part of the radiation image corresponding to this image data S2 is recognized, and image processing is performed on the image data corresponding to this subject part. The image data S3 that has been subjected to image processing is
The radiation image is sent to the image display device 20, and the radiation image is reproduced and displayed based on the image data S3.

FIG. 1 is a diagram showing an example of a radiation image.

The area within the surrounding rectangular frame forms one radiation image 21. This radiographic image 21 includes a subject part 22 which is a radiographic image of the subject (human breast), a direct radiation part 23 in which radiation is directly irradiated to the stimulable phosphor sheet without going through the subject, and only scattered radiation. It can be divided into irradiated scattered radiation portions 24. In the figure, each portion surrounded by a broken line indicates a small region divided such that all the variance values of image data within each portion are equal to or less than a predetermined value.

This division is performed as follows.

First, the entire region of the radiographic image 24 is divided into a plurality of parts such that the area of one small region is x m as shown in the figure.

For each small region having an area lxm divided into a plurality of parts in this manner, a variance value of the image data is determined, and this variance value is compared with a predetermined value. Next, each small area whose variance value is greater than or equal to a predetermined value is further divided into two small areas, and the variance value of each of the divided small areas is determined again and compared with the predetermined value. By repeating the above operations over the entire region of the radiographic image 21 until the variance values of the many divided small regions are all below a predetermined value, for example, the object part 22 and Portions with rapid changes in the radiation image, such as the vicinity of the boundary of the direct radiation section 23, are finely divided.

After being divided as described above, in two adjacent small areas among the many divided small areas, these two
The similarity between the two small areas is evaluated based on the image data corresponding to the two small areas, and when the two small areas are similar, the two small areas are integrated into one small area. Furthermore, similarity evaluation with adjacent small regions is performed, and by repeating this similarity evaluation over the entire region of the radiographic image, the large number of small regions are integrated into a plurality of regions that are dissimilar to each other.

For evaluation of similarity, for example, the average value of image data is calculated for each small area in adjacent small areas, and if the difference between these average values is less than or equal to a predetermined value, the two small areas are determined to be similar. This is done by evaluating whether the

If the accuracy of similarity evaluation is poor only by comparing the average values, the similarity evaluation may be performed using the method shown in FIG. 2.

FIG. 2 is a diagram illustrating one method for evaluating the similarity between two adjacent regions.

Calculate the average value of the image data for each of the two adjacent small areas a and b, and calculate the circumferential lengths A and B of each of the two small areas a and b, and the distance between the two small areas a and b. Find the length X where the above-mentioned average values are in contact with each other, and the value of the difference between the above-mentioned average values is less than or equal to the first predetermined value, and the shorter of the above-mentioned circumference lengths A and B is MIN (A, B)
Then, the value of X/M I N (A, B) is the second
is equal to or greater than a predetermined value, the two small regions a and b are evaluated to be similar. Using this evaluation method, it is possible to improve the accuracy of similarity evaluation compared to a method of simply comparing average values.

Furthermore, it is also possible to obtain a histogram of image data for each small area in two adjacent small areas, and use these two histograms to evaluate the similarity between the two small areas.

FIG. 3 is a graph showing an example of a histogram obtained from a large amount of image data corresponding to each of two adjacent small areas A and B. The horizontal axis of the figure represents a large number of divisions a1+ a2+ . ,...
, and the vertical axis represents each section a1. a2.・・・・・・．

It shows the frequency of appearance of image data, which shows how many pieces of image data exist within each range of ao, .

Between two histograms for each of two adjacent small areas, mutually corresponding sections al, a2, . . .
The absolute value of the difference in appearance frequency for each of . . . , ao, . Δ2,..., Δ.

−1,Δ. , Δn+1.・ is calculated, and the absolute value of this difference is divided into multiple categories a1. a2. .......

ao, · is added, and this added value is compared with a predetermined value, and if the added value is less than or equal to the predetermined value, it is evaluated that the two histograms are similar. In this way, it is evaluated whether two regions are similar or not based on whether two histograms are similar. Also, between the two histograms, the corresponding sections a1+
a2+..., a. 1. Find the difference in appearance frequency for each of the above, find the maximum absolute value of the difference thus found (for example, Δ, -2 in Figure 3), and compare this maximum value with a predetermined value. It may be evaluated whether the two histograms being evaluated are similar.

As exemplified above, the similarity between two adjacent small regions can be evaluated using various methods.

Of course, the above-mentioned similarity evaluation may be performed using any known method other than the above-mentioned example.

As described above, by evaluating the similarity of small regions adjacent to each other over the entire surface of the radiographic image 21 (see FIG. 1), a large number of small regions are integrated into a plurality of regions that are similar to each other.

Thereafter, one or more types of characteristic values are determined for the plurality of regions based on the image data corresponding to each region.

FIG. 4 is a diagram in which the average value and variance value of image data are determined for each of the plurality of regions integrated as described above and plotted on a plane. The horizontal axis shows the average value, and the vertical axis shows the variance value. The position of the O mark in this figure is the number 22, 23, and 24 written inside this Q mark, respectively.
This corresponds to the subject section 22, direct radiation section 23, and scattered radiation section 24 shown in the figure.

In this way, radiographic images of similar subjects under similar imaging conditions are statistically investigated in advance, and the characteristic values that can best separate the subject area from other areas are determined, and the range within which these characteristic values lie. By determining in advance whether the object part is the object part, it becomes possible to recognize the object part from the read image data.

The above embodiment describes a radiation image information reading device that does not perform pre-reading, but it performs pre-reading to obtain pre-read image data, recognizes a subject part based on this pre-read image data, and It goes without saying that the radiation image subject recognition method of the present invention can also be used in a system that determines reading conditions for main reading based on read image data.

Furthermore, this object recognition method is not limited to devices using a stimulable phosphor sheet as shown in FIG.
The present invention can be applied to various devices, such as a device that reads an X-ray image from a film on which the line image is recorded to obtain image data, and reproduces and displays the X-ray image on a display device based on this image data.

(Effects of the Invention) The radiation image subject recognition method of the present invention divides a radiation image into a large number of small regions such that the variance values of image data in each small region are all below a predetermined value, and divides the radiographic image in this way. The similarity of two adjacent small areas among the large number of small areas that have been created is evaluated, and if the two small areas are similar, the two small areas are integrated into one small area and further adjoined. By repeating this similarity evaluation over the entire region of the radiographic image, the large number of small regions described above are integrated into mutually dissimilar regions. Compared to the method of dividing the radiographic image into small regions and integrating them, large regions with little change in the radiographic image are used as large regions for the above similarity evaluation, which saves calculation time. Since it is finely divided, the accuracy of detecting boundaries is improved.

One or more types of characteristic values of the above-mentioned image data are obtained for each of the multiple areas integrated as described above, and the area corresponding to the subject of the radiographic image is recognized based on this characteristic value, so that subsequent image processing In this case, the reading condition 1 image processing condition can be determined so that the area corresponding to this object is reproduced as the best image.

[Brief explanation of the drawing]

Figure 1 shows an example of a radiographic image, Figure 2 shows a method for evaluating the similarity of two adjacent small areas, and Figure 3 shows an example of a histogram. The graph, Fig. 4, is a diagram in which the average value and variance value of image data for multiple small region groups of a radiation image are calculated and plotted on a plane. FIG. 1 is a perspective view showing an example of an image information reading device. 1 - Stimulable phosphor sheet 2.13...Motor 3 Sheet conveying means 4...Ser 6-Rotating polygon mirror 9 Stimulated luminescent light 10...Concentrator 11...Photomultiplier 16.・Amplifier 17...A/D converter 1
8 - Memory 19... Arithmetic unit 20... Image display device 21... Radiation image 22... Subject part 23... Direct radiation part 24 Scattered radiation part = 24- Engraving of drawing ( (No change in content) Figure 1 Figure 2 April 21, 1986

Claims (1)

[Claims] (1) After obtaining a large number of image data carrying radiographic image information by reading a radiographic image in which the subject is recorded in a part, the dispersion value of the image data is calculated for each of a plurality of small regions into which the radiographic image is divided. Each of the small regions whose variance value is equal to or greater than a predetermined value is further divided into a plurality of small regions, the variance value is determined for each of the divided small regions, and finally a large number of divided small regions are obtained. The division is repeated over the entire region of the radiographic image until all the variance values of are equal to or less than the predetermined value. The similarity between these two small areas is evaluated based on the image data corresponding to the area, and when the two small areas are similar, the two small areas are integrated into one small area and further adjoined. By repeating this similarity evaluation over the entire region of the radiation image, the large number of finally divided small regions are integrated into a plurality of regions that are dissimilar to each other. , find one or more types of characteristic values of the image data for each of the plurality of regions,
A method for recognizing a subject in a radiographic image, characterized in that a region corresponding to the subject in the radiographic image is recognized based on the characteristic value.