JP5652883B2 - Contour image generation method used for morphological observation of subject and scoliosis screening system using the same - Google Patents

Description

  The present invention relates to a contour image generation method used for morphological observation of a subject, and a scoliosis screening system using the same,

  An apparatus for determining the surface state of a curved surface of an object using a moire image has been proposed (Patent Document 1). Furthermore, an observation apparatus using a moire topography method is used as an observation apparatus for a human body shape. In such an observation apparatus, fine twists and distortions can be observed as a kind of posture or form observation of an object or a human body by reading an interference pattern (contour line). In addition, it is widely known that the moire topography method is a useful test as a scoliosis screening system.

  An observation device using the moire topography method is a device that projects light onto a target object in a dark place and generates a pattern or striped ripple caused by light, that is, an interference pattern (contour lines). It is an apparatus that generates light and generates contour lines according to the unevenness of the object surface ahead.

JP 7-71934 A

  As described above, an observation apparatus using the moire topography method is useful, but the method using the moire image (the moire topography method) requires special equipment and personnel. In particular, when scoliosis screening is performed using the moire topography method, securing costs and restraint time for screening are problems.

  On the other hand, in recent years, high-resolution digital cameras and portable devices having camera functions have become widespread, and if an image similar to moire topography can be created from these images, it will be inexpensive, time, place and equipment. Scoliosis screening can be realized without any stubbornness.

  Accordingly, an object of the present invention is to easily generate a contour image used for observing the form of a subject from an image photographed by a terminal device having at least a camera function including a general-purpose digital camera or a smartphone, and using this method It is to provide a scoliosis screening system.

  A contour image generation method used for observing the form of a subject that achieves the object of the present invention includes a step of extracting a luminance signal for each pixel from a digital image of a photographed subject, and the extracted luminance signal is used as a predetermined threshold standard. A step of generating a quantized image quantized by the step, a step of converting pixels having the same quantization level for each quantization level in the quantized image, and a pixel having the same quantization level. It has the process of displaying the connected image on a display.

  The contour image generation method for performing scoliosis screening to achieve the object of the present invention includes a step of taking a back image of a subject and outputting a digital image, and a luminance signal for each pixel from the taken digital image. A step of extracting, a step of generating a quantized image obtained by quantizing the extracted luminance signal according to a predetermined threshold standard, and a pixel having the same quantization level are connected for each quantization level in the quantized image. And converting the image into an image, and displaying the image in which the pixels having the same quantization level are connected to each other on a display so that a high degree of deviation on the back surface of the subject can be observed.

  The contour image generation method for performing the scoliosis screening is, as one aspect, that the digital image is in a JPEG image format, and the step of extracting a luminance signal for each pixel from the captured digital image includes the JPEG image A step of converting a format into a BMP image format and a step of converting the BMP image format into a YUV image format, and generating a quantized image obtained by quantizing the extracted luminance signal with a predetermined threshold standard Is characterized in that the Y signal converted into the YUV image format is quantized on the basis of a predetermined threshold.

  According to a second aspect of the contour image generation method for performing scoliosis screening, an image in the YUV image format obtained by the conversion is used as an original YUV image, and a mirror-symmetric image of the original YUV image is used. A step of generating, a step of obtaining a positional relationship in which the two images best overlap the original YUV image and the mirror-symmetric image, and a vertical 2 of the center of gravity of each of the original YUV image and the mirror-symmetric image And a step of using an equipartition line as an approximate symmetry reference, and the left and right inclinations of the body shape of the subject can be determined as the approximate symmetry reference.

  The scoliosis screening system using a contour image used for observing the form of a subject that achieves the object of the present invention includes a portable device having a photographing function and a medical examination center connected to the portable device through a network. Has a photographing function element, a processing device, and a communication function unit, and takes a back image of the subject by the photographing function element, and outputs a luminance signal for each pixel from the photographed digital image by the processing device. Extracting, generating a quantized image obtained by quantizing the extracted luminance signal according to a predetermined threshold standard, and converting the quantized image into an image in which pixels having the same quantization level are connected for each quantization level in the quantized image The image data connecting the pixels having the same quantization level is transmitted to the examination center by the communication function unit, and the examination center is the portable device. An image by connecting the pixels of the same quantization level sent is displayed on the display, characterized by observing the degree of deflection in the back of the subject.

It is a figure which shows the 1st example of a system configuration | structure for performing the contour-line image generation method used for the form observation of the to-be-photographed object according to this invention. It is a flowchart explaining the procedure of the contour-line image production | generation method used by the system of FIG. It is a figure which shows the JPEG image of the to-be-photographed object used as the original image of a JPEG image format. FIG. 6 is a diagram showing a state of a converted YUV image format image represented only by Y (luminance signal). It is a figure explaining the relationship between the distance d from the camera 10 to a to-be-photographed object position, and the luminance signal Y. FIG. It is a figure which shows the example of distribution of the quantized numerical value. It is a figure which shows the image which performed the gradation optimization. It is a figure which shows the contour drawing image of the to-be-drawn subject image. It is a figure which shows the example of an image by a moire type | formula. It is a figure explaining the algorithm which enables a setting with the same unit size as a moire when setting the gradation of a contour line. FIG. 6 is a diagram in which contour lines are drawn on a luminance signal image of a subject image by setting a luminance difference corresponding to 5 mm as a parameter. It is a figure which shows the procedure which calculates | requires a symmetry axis (reference axis) from the contour image (FIG. 8) produced | generated in this invention by approximate symmetry analysis, and sets it as a back midline. It is a figure which shows the expansion Example of this invention. It is a figure which shows the structural example of the scoliosis screening system which performs the process of the contour-line image generation method according to this invention with a smart phone.

  Embodiments of the present invention will be described below with reference to the drawings.

  FIG. 1 is a diagram showing a first system configuration example for executing a contour image generation method used for observing the form of a subject according to the present invention.

  A processing apparatus 1 that processes an image acquired by the digital camera 10 and a display 2 that displays a processing result are provided. The processing apparatus 1 is the same as an ordinary personal computer, and executes the program stored in the memory 14 by the CPU 11 to process the contour image generation method of the present invention.

  Image RAW data acquired by the digital camera 10 or image data equivalent to JPEG is input through an interface (I / F) 16. The input image data is processed by the CPU 11 in accordance with a program stored in the memory 14, and an image in which contour lines are displayed on the subject image is drawn on the video RAM 13. Next, the output of the video RAM 13 is sequentially displayed on the display 2.

  FIG. 2 is a flowchart for explaining the procedure of a contour image generation method used for observing the form of a subject, which is executed by the system shown in FIG.

  In this description, an example of observing a human body shape assuming a scoliosis screening system will be described as an example of morphological observation of a subject.

  First, a subject image is taken by the camera 10 (step S1). RAW data of the subject image photographed by the camera 10 is input to the processing device 1 through the interface 16.

  The input RAW data is converted into a JPEG image format by the CPU 11 (step S2).

  FIG. 3 is a diagram showing a JPEG image of a subject based on the JPEG image format, which is displayed in color. Next, the JPEG image format is converted into bitmap BMP file data in order to make it easier to obtain data for each pixel (step S3).

  Next, the CPU 11 converts the pixel data of the BMP file data into a YUV image format (step S4). The YUV image format is composed of a luminance signal Y, a color difference signal U (B−Y), and a color difference signal V (R−Y). Only Y (luminance signal) is extracted from the YUV image format and stored in the memory 14 (step S5).

  FIG. 4 is a diagram showing a converted YUV image format image, which is represented only by Y (luminance signal). Therefore, it is displayed in black and white monotone.

  Here, the relationship between the luminance signal Y of the photographed image and the distance d from the camera 10 to the subject position will be considered.

  FIG. 5 is a diagram for explaining the relationship between the distance d from the camera 10 to the subject position and the luminance signal Y. FIG. 5A is a schematic diagram illustrating the subject surface from the lateral direction, showing that the distance d from the camera 10 to the subject position varies depending on the unevenness of the subject surface. The distances between the points y1 to y6 on the subject surface and the camera 10 are different, and therefore the magnitude of the luminance signal Y at each point y1 to y6 is smaller in proportion to the distance d from the camera 10. Become. FIG. 5B is a diagram showing the relationship between the distance d and the luminance signal Y by a linear expression (Y = −ad + b).

  Accordingly, when the luminance signal Y of the image is quantized based on a plurality of luminance levels (step S6), the luminance signal image can be represented by a quantized numerical value distribution. FIG. 6 is a diagram illustrating an example of the distribution of the quantized numerical values. It is the figure which quantized the luminance signal of the image which image | photographed the object which has a processus | protrusion in the center part. Since the protruding part is close to the light source, the quantization value becomes large. On the contrary, the numerical value of the recessed part becomes small.

  Here, at the time of quantization, in order to obtain an optimum gradation difference divided into a plurality of stages matching the luminance level, gradation optimization is performed by increasing the contrast. FIG. 7 is an image on which gradation optimization has been performed.

  As described with reference to FIG. 6, the gradation-optimized image is quantized with respect to the luminance value for each pixel. Next, the pixels whose luminance signal level changes are checked, and contour lines are drawn by connecting the pixels (step S7). The obtained contour line data is stored in the memory 14.

  However, with regard to the brightness check, the contour line image becomes a thick line at a location where only the adjacent pixels of the target pixel have changed continuously. Allows identification. In addition, it is possible to draw a contour line that is easy to see by considering a portion where the luminance differs only by one pixel as noise and not drawing.

  Next, the generated contour line image is displayed on the display 2 by converting the contour line data stored in the memory 14 into a JPEG file in an image format (step S8).

  FIG. 8 is a diagram showing a contour drawing image of the subject image drawn in this way.

  Next, the use of the contour drawing image for the scoliosis screening according to the present invention in which the same diagnosis as in the case of the conventional moire image is performed will be described.

  FIG. 9 is a diagram showing an example of an image based on the moire method, and the distance between basic stripes (ring stripes) is set to 5.0 mm, for example. Therefore, considering FIG. 9, the following diagnosis is possible.

  Rings occur in the left and right shoulder blades and buttocks. Since the left and right rings match the left 4th line (L4) and the 7th right line (R7) occurring in the scapula, it can be seen that there are 3 lines on the right side. Therefore, it can be seen that the right side protrudes 5.0 mm × 3 = 15.0 mm (projects forward) from the left side.

  In addition, the inner lower corner of the right scapula has three lines in a slight range. As for this, the same lower angle protrudes extremely to the near side. That is, it represents that the shape is turned to the front side.

  Furthermore, the ring that occurs from the center of the left scapula to the waist is 9 lines (L9), and the right is 12 lines (R12). It can be seen that the left ring appears gently and maintains the shape of the back due to the gentle back muscles. On the other hand, since the density on the right is concentrated at a short distance, the shape is raised from the waist to the right shoulder blade. This means that the right side is stiff, with the quality and shape of the muscle from the lower back to the lower scapular inner corner sticking to the body.

  It will be described that the contour line drawing according to the present invention is possible in the same manner as the image by the moire type.

  FIG. 10 is a device for making the contour line interval correspond to the distance between the basic stripes (ring stripes) in the image example by the moire method being 5.0 mm, for example. That is, FIG. 10 is a diagram for explaining an algorithm that enables setting in the same unit size as that of moire when setting gradations of contour lines.

  The test plate 101 is placed on the wall (background) 100 so that the tip is separated from the wall (background) 100 by 30 mm. The image of the test board 101 photographed with the camera 10 is read, and the brightness values of the upper part and the lower part are acquired. By dividing the acquired value into six equal parts, it is set as a contour contour threshold value. As a result, the distance between the contour lines becomes (30/6 =) 5 mm, which is the same stripe interval as that of the moire image.

  Note that it is only necessary to determine the luminance difference on the image, so setting of the position of the camera 10 from the wall 100 (or subject) is unnecessary.

  That is, the obtained luminance difference corresponding to 5 mm is set as a parameter. When the subject is photographed and analyzed, it is displayed as a contour line for each parameter set by contour drawing with the highest luminance value as a reference. Therefore, irregularities can be grasped at intervals of 5 mm between contour lines in drawing.

  FIG. 11 is a diagram in which contour lines are drawn on the luminance signal image of the subject image by setting the luminance difference corresponding to 5 mm as a parameter as described above.

  From FIG. 11, the following diagnosis is possible.

  The ring occurs in the left and right shoulder blades. Since the left and right rings match the left fourth line (L4) and the right sixth line (R6) occurring in the scapula, 6-4 = 2 and there are two lines on the right. Therefore, the right is raised (projects toward the front) with respect to the left.

  Two lines are generated in a slight range in the lower right corner of the right scapula. As for this, the same lower angle protrudes extremely to the near side. That is, it represents that the shape is turned to the front side.

  The ring extending from the center of the scapula to the waist on the left back is 7 lines, and the right is 15 lines. It can be seen that the left ring appears gently and maintains the shape of the back due to the gentle back muscles. On the other hand, the right side is dense at a short distance, so it has a shape that rises from the lumbar region to the right scapula region. It shows that it has become hard.

  Therefore, the contour image generation method according to the present invention enables a scoliosis diagnostic screening equivalent to the conventional moire method.

  Furthermore, in the diagnosis of scoliosis, it is determined whether the left or right side is inclined with respect to the midline. That is, in the symmetry analysis of the moire image, the midline on the back of the human body is set as a necessary reference axis. The moire image on the back of the subject is almost symmetric with respect to the midline, and can be treated as an approximate symmetric shape.

  Similarly, the symmetry axis (reference axis) is obtained from the contour image (FIG. 8) generated in the present invention by the approximate symmetry analysis, and set as the midline of the back. FIG. 12 shows the procedure.

  12A is an original image corresponding to the contour image in FIG. 8, and FIG. 12B is a mirror-symmetric image obtained by inverting the original image. In each image, center lines A and B passing through the center of the image in the vertical direction are considered.

  Next, an image is created by superimposing these two images. At this time, as shown in FIG. 12 (3), the mirror symmetrical image (FIG. 12 (2)) is moved while being rotated and translated, and the positional relationship (rotation angle) at which the two images best overlap each other. , Translation amount). Then, a vertical line O between the center lines A and B of both images is set as an approximate symmetry reference (midline). Therefore, as in the case of the image based on the moire equation, it is possible to determine the approximate left-right inclination (sideways in the left-right direction) of the body with respect to the approximate symmetry reference (midline).

  FIG. 13 is a diagram showing an extended embodiment of the present invention. An image for each contour line is extracted from the contour line image (FIG. 13 (0)) obtained by the above method (FIG. 13 (1) to FIG. 13 (4)). As a result, it is also possible to display the subject in a circular cut (vertical cut) similar to that obtained by CT scanning or the like.

  That is, in FIG. 13, FIG. 13 (1) corresponds to the state of the re-projection portion being cut off with reference to the maximum threshold level TH 1, and FIGS. 13 (2) to 13 (4) A state in which the subject is cut off corresponding to the levels TH2 to TH4 is shown.

  As described above, the present invention can use a digital camera or a smartphone that has begun to spread widely. It is possible to construct a system for easily screening scoliosis without introducing an expensive measuring device for scoliosis screening.

  Therefore, scoliosis screening can be performed efficiently, and oversight of scoliosis patients can be prevented.

  FIG. 14 is a diagram illustrating a configuration example of a scoliosis screening system in which the process of the contour image generation method according to the present invention is executed by a smartphone.

  The communication function unit 15 of the smartphone 1 as a processing device is connected to the examination center 3.

  As described above, the back image of the subject is photographed by the camera 10 of the smartphone 1 and transmitted to the examination center 3, and the comparison contour image on the back side of the subject is compared with the reference contour image at the examination center 3. Diagnosis is performed by The diagnosis result can be returned from the examination center to the smartphone 1 to notify the subject. In addition, the function of the GPS receiving unit 12 can search for a medical institution near the subject's whereabouts to be detected and send an introduction notification.

  A reference contour image is stored in the memory 14 of the smartphone 1, and the CPU 11 compares the contour image on the back of the subject photographed by the camera 10 with the reference contour image stored in the memory 14. The smartphone 1 itself can be configured to determine the possibility of scoliosis.

  As described above, according to the present invention, instead of the conventional moiré-type object shape observation or scoliosis screening, it is possible to realize an object shape observation or scoliosis screening system with a simple apparatus configuration and labor saving. It is.

DESCRIPTION OF SYMBOLS 1 Processing apparatus, smart phone 2 Display 3 Examination center 10 Camera 11 CPU
12 GPS receiver 13 Video RAM
14 Memory 15 Communication Function Unit

Claims (5)

A contour image generation method for scoliosis determination,
Taking a back image of the subject and outputting a digital image;
Extracting a luminance signal for each pixel from the photographed digital image;
Generating a quantized image obtained by quantizing the extracted luminance signal with a predetermined threshold standard;
For each quantization level in the quantized image, converting the pixel of the same quantization level into an image connected,
Displaying an image connecting pixels of the same quantization level on a display to enable observation of a high degree of deviation in the back of the subject;
The test plate is preliminarily arranged obliquely so that the depth between the front end and the rear end has a predetermined interval, and the luminance value of the front end and the rear end of the test plate to be imaged is acquired. Is a value obtained by equally dividing a predetermined number into the predetermined threshold reference in the quantization,
A contour image generation method for determining scoliosis characterized by the above. In claim 1,
The digital image is in a JPEG image format, and the step of extracting a luminance signal for each pixel from the photographed digital image includes:
Converting the JPEG image format to a BMP image format;
Converting the BMP image format to a YUV image format,
The step of generating a quantized image obtained by quantizing the extracted luminance signal with a predetermined threshold standard quantizes the Y signal converted into the YUV image format with a predetermined threshold standard.
A contour image generation method for determining scoliosis characterized by the above. A scoliosis screening system using a back image of a subject,
An imaging function element to shoot the back image of the subject,
Having a processing device,
The processor is
A luminance signal extracting means for extracting a luminance signal for each pixel from a digital image photographed by the photographing functional element ;
A quantized image generating means for generating a quantized image obtained by quantizing the extracted luminance signal with a predetermined threshold reference;
Wherein the quantization level for each of the quantized image, to function as a image converting means for converting the connected images pixels of the same quantization level, further,
A display that observable displaying a high degree of deflection of the image by connecting the pixels of the same quantization level in the back of the front Symbol subject,
Depth-edge and trailing edge has a test plate disposed obliquely so as to have a predetermined interval,
Furthermore, the imaging functional element, when taking a back image of the subject, previously image the test board,
The luminance signal extraction unit of the processing device acquires the luminance values of the front end and the rear end of the test plate, and the quantized image generation unit is obtained by equally dividing the acquired luminance value by a predetermined number. A value is obtained, and the obtained value is used as the predetermined threshold reference in the quantization,
A scoliosis screening system characterized by that. In claim 3,
The digital image is a JPEG image format, and the luminance signal extraction means for extracting a luminance signal for each pixel from the photographed digital image,
Converting the JPEG image format to a BMP image format, converting the BMP image format to a YUV image format ,
Wherein the quantized image generating means for the extracted luminance signal to generate a quantized image obtained by quantizing a predetermined threshold criterion, to Y signal converted into the YUV image format, quantized with predetermined threshold criteria To
A scoliosis screening system characterized by that. In claim 4,
The quantized image generating means includes
An image in the YUV image format obtained by the conversion is used as an original YUV image, and a mirror-symmetric image of the original YUV image is generated .
The mirror Film determined Me a positional relationship that the symmetrical image is the two images overlap best with the original YUV image,
In the determined positional relationship, an intermediate vertical line between vertical center lines passing through the centers of the original YUV image and the mirror-symmetric image is used as an approximate symmetry reference, and
The processor is
The approximate symmetry criteria as midline of the back of the subject, based on the median line, which functions as a means for further determining a lateral inclination of type of the subject,
A scoliosis screening system characterized by that.