Image processing method

Abstract

Provided is an image processing method for visualizing fine irregularities in a lower layer portion covered by blood. An original image is converted into an RGB color space, filter monochromatization is performed on each pixel of the image converted into the RGB color space, a gray plane is generated, and when the gray plane is generated, a color emphasis filter is used, and the luminance ratio (Y1) of only a region having a specific hue is increased. Further, the original image is converted into a YUV color space, a YCbCr color space, or a YPbPr color space, the converted image and the gray plane are compared for each pixel by luminance (Y), the higher (bright) one is used as a color image, and the color image is sharpened.

Description

Technical Field

[0001] The present invention relates to an image processing method for visualizing fine irregularities of a lower layer portion covered with blood, for example. Background Art

[0002] The brightness and contrast of the image are emphasized according to the purpose of use, etc. For example, Patent Document 1 discloses an image processing device for the purpose of finding blood vessels or bleeding sites during endoscopic surgery.

[0003] Specifically, disclosed are: an image acquisition unit that acquires a camera image including a subject image obtained by irradiating illumination light from a light source unit to the subject; a visual recognition emphasis unit that performs processing on areas other than yellow in the camera image so as to relatively improve the visual recognition of the yellow area in the camera image; a blood area detection unit that detects an area of ​​blood in the camera image based on color information of the camera image; and a notification processing unit that performs notification processing on the blood area based on the detection result of the detection unit.

[0004] In Patent Document 2, as adjustment items, brightness, chroma, red emphasis, skin color correction, contrast, and scratch noise addition can be listed. As an example of setting red emphasis, an image obtained by photographing red leaves can be listed. As an example of brightness enhancement, an image obtained by photographing the aurora can be listed.

[0005] Patent Document 3 discloses, as a method of imaging a vascular system, a highlighted display method in which veins are colored red and arteries are colored blue to distinguish between pixels representing venous vascular systems and pixels representing arterial vascular systems, and the veins are highlighted with different highlights.

[0006] Patent document 4 discloses an image processing method for endoscopic images. The image processing method discloses dividing an image into a first region where the oxygen saturation exceeds a fixed value and a second region where the oxygen saturation is lower than the fixed value, performing color balancing processing on the first region, performing gain processing for changing the color of the second region, and performing red emphasis processing for emphasizing the red component as the color balancing processing.

[0007] Patent Document 5 discloses a method of easily identifying a bleeding site and finding a fine bleeding site by using an optical lens having polarization properties for emphasizing red color and provided integrally with a polarization filter.

[0008] Prior Art Literature

[0009] Patent Literature

[0010] Patent Document 1: WO2018-235179

[0011] Patent Document 2: Japanese Patent Application Publication No. 2020-99014

[0012] Patent Document 3: Japanese Patent Application No. 2020-508713

[0013] Patent Document 4: Japanese Patent Application Publication No. 2014-50453

[0014] Patent Document 5: WO2020-031502 Summary of the invention

[0015] Problems to be solved by the invention

[0016] Patent document 1 discloses a process for enhancing the chromaticity and brightness of the color of the blood vessel area. However, this process is for making the bleeding site and the location of the blood vessel easier to understand. Unlike the present application, the process for making the lower part (the lower layer covered by the blood) that is difficult to observe due to the blood clear is not disclosed, that is, the process for increasing the contrast by expanding the gradient of the red area to make the red thinner is not disclosed.

[0017] Patent Documents 2 to 5 also relate to processing of making the red portion conspicuous relative to other portions, rather than visualizing the image of the portion under the blood that is hidden by the blood (red portion).

[0018] That is, the present invention does not emphasize the processing of red, but aims at processing to eliminate red, and this point is not disclosed in any patent document.

[0019] Means for solving problems

[0020] The present invention does not make a specific color, such as red, more conspicuous than other parts, but removes the adverse effects of red and makes the parts below it visible. To this end, the image processing method involved in the present invention transforms the original image into an RGB color space, performs filtering and monochromatization on each pixel of the image transformed into the RGB color space, generates a gray plane, and uses a color emphasis filter when generating the gray plane to increase only the brightness rate (Y1) of the area with a specific hue. Further, the original image is transformed into a YUV color space, a YCbCr color space, or a YPbPr color space, and the brightness (Y) is used to compare the transformed image and the gray plane for each pixel, and the higher (brighter) one is used as the color image, and the color image is sharpened.

[0021] When the purpose is to visualize fine irregularities of the lower layer portion covered with blood, a red emphasis filter is used as a color emphasis filter to increase the brightness rate (Y1) of only a region having a red hue close to that of blood.

[0022] In addition, it is considered that the color emphasis filter emphasizes a specific color by increasing the brightness of the specific color to make it close to white.

[0023] Effects of the Invention

[0024] According to the image processing method involved in the present invention, for example, a portion covered by blood can be visualized during surgery, so that surgery can be performed while confirming the location of smaller bumps and wounds, thereby improving safety. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 It is the original image with bleeding.

[0026] Figure 2 It is a black and white image of the original image with bleed.

[0027] Figure 3 This is a black and white image with the red area of ​​the original image with bleeding enhanced in brightness.

[0028] Figure 4 This is a color image obtained by synthesizing the replaced luminance signal (Y1), the difference between the luminance signal and the blue component (U), and the difference between the luminance signal and the red component (V).

[0029] Figure 5 This is an image obtained by sharpening the synthesized color image. DETAILED DESCRIPTION

[0030] Hereinafter, embodiments of the present invention will be described based on the drawings.

[0031] Figure 1 This is an original image with bleeding. From this image, it can be seen that the condition of the portion covered by the blood cannot be observed.

[0032] Figure 2 The original image is converted to RGB color space and filtered to a black-and-white image. The red color is high but the brightness is reduced and dark. In other words, it can be seen that the part covered by blood cannot be visualized because the brightness is reduced.

[0033] For this reason, Figure 3 As shown, a black-and-white image is created by brightening a region (bleeding region) having a red hue in the original image using a red emphasis filter (brightness ratio: Y1).

[0034] In the filtering and monochromatization process, the weights are determined for each hue of R (red), Ye (yellow), G (green), C (cyan), B (blue), and M (magenta) from the RGB image, and the image is converted into grayscale.

[0035] If the weights are set to Rf (red), Yef (yellow), Gf (green), Cf (cyan), Bf (blue), and Mf (magenta) for each hue, then in the present embodiment, Rf = 100%, Yef = 60%, Gf = 40%, Cf = 60%, Bf = 20%, and Mf = 80%. In addition, the brightness sensitivity of humans to red expressed by the hue circle is Rf = 40%.

[0036] By specifying a negative ratio in the above f, it is also possible to weaken a specific color.

[0037] Assuming that the black and white value (brightness ratio) of a pixel having R, G, and B values ​​is Y, the following calculation is performed.

[0038] Here, max=the maximum value of RGB, mid=the value that is neither the maximum nor the minimum of RGB, and min=the minimum value of RGB. However, in the case of having the same value, mid becomes the same value as max or min.

[0039] (1) When max==mid==min, set the value to Y and the calculation ends.

[0040] (2) Starting from Y = 0

[0041] (3) When max==R, Y+=Rf·(max-mid)

[0042] (4) When max==B, Y+=Ye f·(max-mid)

[0043] (5) When max==G, Y+=Gf·(max-mid)

[0044] (6) When min==R, Y+=Cf·(max-mid)

[0045] (7) When min==B, Y+=Bf·(max-mid)

[0046] (8) When min==G, Y+=Mf·(max-mid)

[0047] (9) There must be two cases from (3) to (8) that match (three cases with the same value)

[0048] (10) Y+ = min

[0049] (11) If Y>255, then Y=255

[0050] (12) If Y < 0, then Y = 0

[0051] For example, the brightness of red represented by the hue ring is 40%, but this brightness is multiplied by 2.5 to 100%. By converting the brightness of red to 100%, the red area in the original image, which is composed of a gradient from black to red, is converted to a wide gradient from black to red to white. It can be seen that by converting the gradient to a wider red area, the red area is Figure 2 In comparison, even in the same black and white photo, subtle differences in light and dark are visualized.

[0052] In the present invention, the magnification of the brightness ratio of red is an adjustable parameter and is not limited to 2.5 times, but may be 2.0 times or the like.

[0053] In the case of veins with a large amount of blue components, not only R (red) but also M (magenta) magnification is changed, so that the difference in color tone of the blood image can be adjusted. By setting the magnification corresponding to the color tone of the magnification for R (red), Ye (yellow), G (green), C (cyan), B (blue), and M (magenta), an area with any color tone can be transmitted.

[0054] Next, the brightness ratio (Y1) of the red area of ​​the black-and-white image brightened by the red enhancement filter is replaced with the brightness ratio (Y) of the red area of ​​the original image in the YUV color space, YCbCr color space, or YPbPr color space. That is, the brightness ratio (Y) of the red area of ​​the original image is set to 2.5 times. The color image produced by synthesizing the replaced brightness ratio (Y) with the UV, CbCr, or PbPr of the original image is shown in FIG. Figure 4 .

[0055] like Figure 4 As shown, by increasing the brightness of the red area, the dark part in the original image becomes brighter, the gradation becomes wider, and the unevenness of the part below the blood is visualized. However, at the same time, the red becomes a fluorescent color at the boundary of the red area.

[0056] The reason for the above is that when the image is decomposed into YUV and Y (brightness) is passed through a red enhancement filter to brighten only the red area (for example, 2.5 times), the area that was originally black to dark red has a wide color gradation from black to pink beyond red and into white, and the appearance becomes a fluorescent color image mixed with bright pink, resulting in the omission of smaller bumps and scratches.

[0057] In order to eliminate the above phenomenon, the widened gradation is returned to a natural brightness range. Specifically, by using PN.6867563, PN.6893068, CLAHE, LISr (registered trademark), etc., a sharpening process is performed to adaptively make the contrast uniform.

[0058] Figure 5This is a color image after the sharpening process. It can be seen that the unevenness of the portion covered by blood is more visible than in the original image.

[0059] Industrial Applicability

[0060] In the embodiment, the portion covered by blood during surgery is visualized by increasing the brightness rate of the red region and expanding the gradient of the red region, but the brightness rate of regions other than red can also be increased. In particular, it can be used to clarify the lower layer portion of the image region with high chromaticity but low brightness.

Claims

1. An image processing method, comprising the following steps: (1) Transform the original image into RGB color space; (2) Determine the weights of the hues of red (R), yellow (Y), green (G), cyan (C), blue (B), and magenta (M) from the image transformed into the RGB color space and transform it into grayscale; (3) when transforming to the grayscale, using a red emphasis filter to increase the brightness rate Y1 of the red area of ​​the black and white image; (4) Convert the original image to YUV color space, YCbCr color space or YPbPr color space; (5) replacing the brightness of the red area of ​​the black-and-white image brightened by the red emphasis filter with the brightness Y of the red area of ​​the original image in the YUV color space, YCbCr color space, or YPbPr color space; and (6) In order to return the gradation that has been widened by the above-mentioned processing to a natural brightness range, the contrast is made uniform.

Images