A high-precision image sub-region coloring display method

Abstract

The application discloses a high-precision image sub-region color display method, and belongs to the field of image processing and display, specifically, a method of integrating multiple groups of preset window levels and window widths is used to transform input image data, and the transformed image is colored with different colors according to preset window levels and window widths, so that the image content is completely displayed in sub-regions; in order to make the visual system resolution of an observer reach the maximum, a region-following color method is used to dynamically adjust the brightness value of a pixel at a focus point in an observation region of the observer; in order to make the observer effectively distinguish image details and correctly identify image contours, a region-level difference modulation technology is used to adjust the gray level difference between image pixels in the observation region. The application solves the defect that the existing window display technology easily covers part of the image with specific gray values, does not need to specially divide regions, and can effectively improve the efficiency of a display system.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of image processing and display, and particularly relates to a high-precision image sub-region coloring display method. BACKGROUND

[0002] With the development of science and technology, the image precision produced by nondestructive testing technology is higher and higher, and the pixel data length is longer and longer. The image display is generally divided into hard display and soft display. The hard display is the image presentation on a printing paper or a film medium, and the soft display is the image display on a display through software.

[0003] For the hard display of the image, specific chemicals are deposited on the film according to the image gray scale through chemical reaction. These chemicals reflect different amounts of light to the human eye under the irradiation of the light source, thereby forming the image. When a high-precision image needs to be displayed on a low gray scale medium, the windowing technology of the soft display of the image is used to select the image presentation form under specific parameters, and then printing or printing is performed to solidify the image on the corresponding medium.

[0004] For the soft display of the image, the screen pixels emit light in the same proportion as the image to the human eye, thereby forming the image. When the range of the image display device is smaller than the gray scale range of the image, the windowing display technology is used to display the high-precision image on the display device.

[0005] The windowing display technology selects the data range (VOI, Value of Interest) to be processed according to the original image data, and the VOI is generally obtained by adjusting the window width and the window level according to the habit of the observer. In the specific VOI range, finding the region of interest (ROI) is the final purpose of the observer. The so-called ROI has different definitions in different fields. In the field of exploration and device detection, the ROI generally corresponds to faults, cracks, etc., and in the medical field, the ROI generally corresponds to the lesion tissue.

[0006] The window width and the window level used by the windowing display technology specify a linear change between the original pixel value and the display value. Assuming that the image data length is N bits, the image data length supported by the display device is M bits (N>M), and the windowing display technology maps the 2 N level gray scale to the 2 M level gray scale interval, thereby realizing the display of the high-precision image.

[0007] In order to facilitate the observer to find the ROI area, a preset window level and window width method is generally adopted to realize the display of the image of a specific area. Since the image gray data range corresponding to different preset data is not completely the same, and different image gray data ranges correspond to the image content of different areas in the image data. Therefore, the preset window level and window width method can only display the image content of the corresponding area in the data range, and cannot display all the image data at the same time.

[0008] Therefore, in actual operation, the window display technology has the disadvantage of easily covering part of the image of a specific gray value, specifically:

[0009] 1. When the selected window level C and window width W are selected, only the image with a gray value of C-W / 2 to C+W / 2 in the image can be displayed, and the image content of other areas cannot be normally displayed;

[0010] 2. When W » 2 M , the multiple gray values of the input image pixel data can only be mapped to the same gray value on the display medium, and the image content with a smaller gray value range cannot be displayed;

[0011] 3. When the image content is normally represented, the human visual system has difficulty in identifying the image details in the visual attention area when the image pixel gray range is greater than 20-30. SUMMARY

[0012] The present application provides a high-precision image sub-regional coloring display method to solve the defects in the above-mentioned image soft display technology, and realizes full display of the gray or brightness of high-precision images during image soft display, so as to facilitate the observer to find the ROI area.

[0013] The high-precision image sub-regional coloring display method has the following specific steps:

[0014] First, input the image data to be displayed, transform it through a transformation function, and sequentially color each pixel after transformation according to the preset window level and window width to realize a new completely displayed image G(u, v) with distinction;

[0015] The transformation function is as follows:

[0016]

[0017] a1=(2 M -1) / W1, b1=(2 M -1)(1 / 2-C1 / W1)

[0018] a2=(2 M -1) / W2, b2=(2 M -1)(1 / 2-C2 / W2)

[0019] …

[0020] a j =(2 M -1) / W j , b j =(2 M -1)(1 / 2-C j / W j )

[0021] Wherein, C j is the window position of the jth preset window position; W j is the window width value of the jth window width, S j indicates the color corresponding to the pixel of the preset window position and window width sequence value j; a j =(2 M -1) / W j , b j =(2 M -1)(1 / 2-C j / W j ), and C j -W j / 2≤f(u,v)≤C j +W j / 2; M is the length of the pixel data in the image supported by the display device;

[0022] The formula indicates that if the input image pixel gray scale is f(u,v), the monochrome with the brightness value a j f(u,v)+b j is outputted, and the specific color is determined by the sequence of the preset window position and window width;

[0023] Then, the image content of different regions in the image corresponding to different preset window positions and window widths is integrated to display all the image content in the image data.

[0024] Finally, the VOI region needing to be processed is selected according to the original image data, the brightness value of the pixel at the focus point in the observer's attention region is adjusted by using the follow-up coloring method, so that the resolution of the observer's visual system reaches the maximum; at the same time, the step difference between the gray scales of the detail pixels in the attention region is adjusted by using the regional step difference modulation method, so that the observer can effectively distinguish the image details or correctly identify the image contour.

[0025] The advantages of the present application are:

[0026] The application discloses a high-precision image sub-regional color display method, which does not need to perform special region division and can effectively improve the efficiency of a display system; a region inner level difference modulation technology is used, so that when the image gray scale difference in the region is greater than the image gray scale difference that can be presented by a display medium, the image details can be effectively distinguished; a region inner follow-up color technology is used, so that when the image pixel gray scale in the adjacent region boundary field is similar or the same, the image details can be recognized. BRIEF DESCRIPTION OF DRAWINGS

[0027] Figure 1 is a flow chart of the high-precision image sub-regional color display method of the application;

[0028] Figure 2 is a simulation monochrome image curved surface schematic diagram of the application.

[0029] Figure 3 is a transformation curve corresponding to a plurality of groups of window position and window width preset data of the application.

[0030] Figure 4 is a transformation curve integrated by a plurality of groups of window position and window width preset data of the method of the application.

[0031] Figure 5 is a sub-regional color effect diagram of the method of the application.

[0032] Figure 6 is a sub-regional color transformation schematic diagram when N=5 and M=3 of the application.

[0033] Figure 7 is a sub-regional color transformation schematic diagram when N=5 and M=3 of the application, in which the overlapping region is transformed according to the left curve.

[0034] Figure 8 is a sub-regional color transformation schematic diagram when N=5 and M=3 of the application, in which the overlapping region is transformed according to the right curve.

[0035] Figure 9 is a sub-regional color transformation schematic diagram when N=5 and M=3 of the application, in which the overlapping region is transformed according to the priority of the concerned region.

[0036] Figure 10 is a sub-regional color transformation schematic diagram when N=5 and M=3 of the application, in which the region inner follow-up color is used (the curve is moved to the left).

[0037] Figure 11 is a sub-regional color transformation schematic diagram when N=5 and M=3 of the application, in which the region inner follow-up color is used (the curve is moved to the right).

[0038] Figure 12 is a sub-regional color transformation schematic diagram when N=5 and M=3 of the application, in which the region inner image level difference modulation technology is used (the level difference is increased).

[0039] Figure 13is a schematic diagram of image level difference modulation technology (smaller level difference) in the area of N=5, M=3 of the present application. DETAILED DESCRIPTION

[0040] Before the embodiments are described in detail, the technical solutions of the present application are described in detail in conjunction with the drawings. In the following description, several different aspects of the present application will be described, however, the present application can be implemented only by using some or all of the structures or processes of the present application for those skilled in the art. In order to clarify the explanation, specific numbers, configurations and sequences are described, but it is obvious that the present application can be implemented without these specific details. In other cases, in order not to confuse the present application, some well-known features will not be described in detail.

[0041] The present application is a high-precision image sub-region coloring display method, as shown in the figure, the specific steps are as follows: Figure 1

[0042] First, input the image data to be displayed, transform it through the transformation function, and color each pixel after transformation according to the preset window level and window width in order to realize a new completely displayed image G(u, v) with different colors.

[0043] The transformation function is as follows:

[0044]

[0045] a1 = (2 M -1) / W1, b1 = (2 M -1)(1 / 2-C1 / W1)

[0046] a2 = (2 M -1) / W2, b2 = (2 M -1)(1 / 2-C2 / W2)

[0047] …

[0048] a j =(2 M -1) / W j , b j =(2 M -1)(1 / 2-C j / W j )

[0049] Wherein, C j is the window level of the jth preset window level; W j is the window width value of the jth window width; S j represents the color corresponding to the pixel with the sequence value j of the preset window level and window width; a j =(2 M ​-1) / W j b j =(2 M -1)(1 / 2-C j / W j ), and satisfy C j -W j / 2≤f(u,v)≤C j +W j / 2; M is the length of pixel data in the image supported by the display device;

[0050] This formula indicates that if the grayscale value of the input image pixel is f(u,v), then the brightness value is a. j f(u,v)+b j The specific color output is determined by the preset window level and window width values.

[0051] like Figure 2 As shown, in the continuous domain, the brightness value of a simulated grayscale image is represented as f(x,y). When the grayscale value range of f(x,y) after quantization is larger than the grayscale value range that the display medium can display, the image cannot be fully displayed. However, when taking a small region S of the image, if the grayscale value difference range P of the image within that region is smaller than the grayscale value range 2 that the display medium can display, the image will not be fully displayed. M (M is the number of bits that the display medium can display). Then, the gray values ​​of all image pixels in the area are mapped to the gray value range that the display medium can display in a linear relationship, and the image in the area can be displayed normally.

[0052] Then, by integrating multiple sets of preset window levels and widths, the transformation curves corresponding to all preset window levels and widths are integrated together, and the input image data is transformed using a transformation function.

[0053] Since different preset window positions and window widths correspond to different areas of image content in the image, integrating multiple preset window positions and window widths can display all image content in the image data.

[0054] Windowed display technology can display specific image content by pre-setting window positions and widths corresponding to different ranges of image grayscale data. These pre-set window positions and widths generally correspond to different areas of image content. The region-based coloring technology provided by this invention integrates the transformation curves corresponding to the pre-set window positions and widths, uses a transformation function to transform the input image data, and applies different colors to the image content according to the pre-set window positions and widths. Due to the correlation of image pixels—that is, adjacent or nearby pixels have the same or similar grayscale or color—after the input image is transformed by the formula and then displayed, many adjacent but different colored regions will be formed, such as… Figure 5As shown. At this point, the images of different regions are distinguished by color, and the pixels within a region are distinguished by color brightness. The observer can also adjust the color scheme of each region, but should follow the rule that regions corresponding to different transformation curves should be colored with different colors, while different regions corresponding to the same transformation curve should be colored with the same color.

[0055] If the transformation curves corresponding to adjacent preset window positions and window widths overlap, such as Figure 7 As shown, the transformation of this portion of the grayscale image data can be uniformly defined as being processed according to the rule of transformation based on the left or right curve. When observing, the observer can adopt the principle of prioritizing the area of ​​interest, that is, if the transformation curve of the area where the observer's point of interest is located overlaps with the transformation curve of the adjacent area, then the grayscale of the overlapping part of the image data will be transformed according to the transformation curve of the area where the point of interest is located.

[0056] Finally, based on the original image data, the VOI region to be processed is selected, and the brightness value of the pixels at the point of interest in the region of interest is adjusted dynamically using the follow-up coloring method to maximize the resolution of the observer's visual system. At the same time, the level difference modulation method within the region is used to adjust the level difference between the gray values ​​of the detail pixels in the region of interest, so that the observer can effectively distinguish the image details or correctly identify the image content.

[0057] Within the observer's area of ​​interest, the brightness value of the pixel at the point of interest within the area of ​​interest is dynamically adjusted using the intra-regional coloring method, so that the observer's visual system resolution is maximized and image details are better distinguished. Within the observer's area of ​​interest, the intra-regional differential modulation technique is used to adjust the level difference between the gray values ​​of the detail pixels within the area of ​​interest, so that the observer can effectively distinguish image details or correctly identify image content.

[0058] After transformation using the formula, a monochrome brightness level less than or equal to that of the display medium within a monochrome region can represent a grayscale level, allowing the observer to normally locate the Region of Interest (ROI) within the monochrome region. However, due to the limitations of the human visual system in terms of resolution—that is, when the image brightness is too strong or too weak, or when the background brightness is too strong—the human eye's ability to distinguish images decreases. Therefore, if the image pixels at the point of interest selected by the observer within the monochrome region are brighter or darker than the selected point of interest, the human eye will have difficulty distinguishing that image, and thus cannot quickly determine the ROI region.

[0059] To improve this situation and enable the observer to better focus on the monochromatic region S i By finding the ROI region, the formula transformation can be made dynamic.

[0060] The specific method is as follows: After the formula transformation and display, the observer selects a monochrome area. If the coordinates of the point of interest are (u, v), the brightness value of the color after the grayscale value of the image pixel at that point is transformed is a. i f(u,v)+b iIf the luminance value is relatively bright or dark to the observer, the luminance value of a certain pixel at this point is taken as the middle value 2 of the luminance value that the image display can present M-1 .

[0061] If a i f(u, v) + b i < 2 M-1 , let A i = 2 M-1 - a i f(u, v) - b i , f A (u, v) = f(u, v) - (A i - b i ) / a i .

[0062] Then f A (u, v) is substituted into the corresponding i-th transformation formula, and the luminance of all the pixel points in the area where the observer's focus point is located is transformed, while the luminance of the pixel points in other areas does not change.

[0063] If a i f(u, v) + b i > 2 M-1 , let B i = a i f(u, v) + b i - 2 M-1 , f B (u, v) = f(u, v) + (B i - b i ) / a i . Then f B (u, v) is substituted into the corresponding i-th transformation formula, and the luminance of all the pixel points in the area where the observer's focus point is located is transformed, while the luminance of the pixel points in other areas does not change.

[0064] When the observer's focus point changes again, the above process is repeated. This transformation process appears on the transformation curve as follows: the transformation curve segment corresponding to the i-th transformation formula where the focus point is located moves along the horizontal axis, while other transformation curve segments do not move.

[0065] After such processing, the luminance range of the image presented by the display medium in the observer's focus area is always within the best range of the visual system's image resolution capability, so the observer's visual system has the strongest ability to distinguish image details. However, the following adverse consequences are likely to occur: part of the image data of the gray scale does not correspond to the transformation curve, which is likely to form a blind area of observation; and another part of the image data of the gray scale corresponds to two transformation curves, such as Figure 4At this time, the image data gray scale of the blind area part can be displayed by automatically setting the auxiliary conversion curve, and the gray scale conversion of the image data of the overlapping part can be realized by the principle of focusing on the key points.

[0066] When the level difference between the gray scale values of the detail pixels in the area of interest of the observer is small, the observer can not be able to distinguish the image details. Or when the level difference between the gray scale values of the image pixels in the area is large, the observer can not be able to distinguish the image content. At this time, the level difference modulation technology in the area can be used to increase or decrease the level difference of the image pixel gray scale values, so as to realize the resolution of the image details or the recognition of the image contour.

[0067] The specific use method of the method is: taking the brightness of the key point pixel in the current area of interest as the window level, and then adjusting the window width value corresponding to the conversion curve corresponding to the color of the region. If the window width value is small, the color brightness difference of the pixels with similar gray scale values in the original image data will be large after being converted into the corresponding color, which is beneficial to the observer to distinguish the image details with similar brightness values; and if the window width value is large, the color brightness difference of the pixels with similar gray scale values in the original image data will be small or become the same color, which is beneficial to the observer to observe the overall contour of the image.

[0068] Embodiment 1

[0069] Suppose that an N=5-bit image will be displayed on an M=3-bit gray scale display, that is, an image with 32 gray scales from 0 to 31 is displayed on a display supporting 8 levels of color brightness from 0 to 7. Under normal circumstances, the image cannot be completely displayed. If the technology of the present application is used, the image will be divided into different color regions in the display, and the number of color levels in each region is less than or equal to the number of color levels supported by the display. For example Figure 6 The curve set composed of four different window level and window width preset value data is integrated together in the middle, and the function curve of the upper part is formed. Figure 6 According to the curve, the input image data is converted, and the converted image is colored with different colors according to the preset window level and window width, so that all the image data can be displayed at the same time.

[0070] Figures 7-13 In the middle, the data box in the middle represents the different color brightness levels supported by the display, from 0 to 7, a total of 8 levels of brightness; the lower data box represents the gray scale values of the input image, from 0 to 31, a total of 32 levels of gray scale. Take a pixel with any gray scale value from the image, and the converted color can be obtained through the conversion curve. For example Figure 5 In the middle, the pixel with a gray scale value of 17, that is, Figure 6 17 in the lower data box in the middle, is colored according to Figure 6The image pixel gray value 17 will be displayed as S2 color with brightness 6 in the display medium.

[0071] Embodiment 2

[0072] If the integrated transform curves have overlapping areas, such as the adjacent areas of S1, S2 and S3 as shown in Fig. 2, the image data transform of the overlapping part can be uniformly defined as being processed according to the rule of left curve transform or right curve transform. Figure 7 The example in Fig. 2 is processed according to the rule of left curve transform, i.e. in the transform, the image data of the overlapping part is uniformly transformed according to the left curve of the adjacent curves in the integrated transform curves. For example, the pixel with gray value 17 in Fig. 1, i.e. 17 in the lower data frame of Fig. 2, is transformed according to the left curve of the transform curves in Fig. 1, corresponding to 7 of S2 color in the middle data frame, instead of S3 color in the middle data frame. In the display, the image pixel gray value 17 will be displayed as S2 color with brightness 7 in the display medium. Figure 7 Figure 7 Figure 7 Figure 7 Figure 8 The example in Fig. 3 is processed according to the rule of right curve transform, i.e. in the transform, the image data of the overlapping part is uniformly transformed according to the right curve of the adjacent curves in the integrated transform curves. In the display, the image pixel gray value 17 will be displayed as S3 color with brightness 2 in the display medium.

[0073] Embodiment 3

[0074] If the integrated transform curves have blind areas, such as the adjacent areas of S4 and S5 as shown in Fig. 4 and the adjacent areas of S3 and S4 as shown in Fig. 5, the method of fixed color can be used for processing, as shown in Fig. 6; or the display can be realized through the auxiliary transform curves set in advance, as shown in Fig. 7. The auxiliary transform curves set in advance can be in various forms, such as the auxiliary transform curves in Fig. 7 are formed by the original transform curves (shown by dotted lines) moving in a cycle. Figure 10 Figure 11 Figure 10 Figure 11 Figure 11 The auxiliary transform curves in Fig. 7 are formed by the original transform curves (shown by dotted lines) moving in a cycle.

[0075] Embodiment 4

[0076] When the observer observes, the principle of attention area priority can also be used, i.e. if the transform curves of the area where the observer's attention point is located and the adjacent area overlap, the gray of the image data of the overlapping part is transformed according to the transform curve of the area where the observer's attention point is located. For example, as shown in Fig. 8, if the area where the observer's current attention point is located is S2 color, all the image pixels in this area are transformed according to the transform curve corresponding to S2 color. Figure 9

[0077] ​​​​​​​​​After the initial conversion, if the observer is interested in a certain monochrome image area, moving the conversion curve corresponding to the color along the horizontal axis left and right can achieve the effect of following the color in the area. If the area corresponds to S4 color, the pixel color brightness value of the point of interest is 1, as shown in FIG. 8. At this time, moving the conversion curve to the right makes the pixel color brightness value of the point of interest change to 3 or 4, which is stronger in the resolution ability of the human eye, as shown in FIG. 9. Figure 9 If the area corresponds to S4 color, the pixel color brightness value of the point of interest is 6, as shown in FIG. 10. At this time, moving the conversion curve to the left makes the pixel color brightness value of the point of interest change to 3 or 4, which is stronger in the resolution ability of the human eye, as shown in FIG. 11. Figure 11 Figure 9 Figure 10 Because the left and right movement of the data corresponding to the conversion curve in the area will cause overlapping areas and blind areas in the data conversion in the area, the method in Embodiments 2 and 3 is used for processing.

[0078] Embodiment 5

[0079] If the image level difference modulation technology in the area is to be realized, the brightness of the pixel of the point of interest in the current area of interest is taken as the window level, and then the window width value corresponding to the conversion curve corresponding to the color in the area is adjusted separately. As shown in FIG. 12, if the observer's current point of interest is the image pixel with a brightness of 3 in S4 color, the window width value is reduced, and the resolution of the image details can be realized. As shown in FIG. 13, if the observer's current point of interest is the image pixel with a brightness of 3 in S4 color, the window width value is increased, and the recognition of the image contour can be realized. Figure 12 Figure 13

[0080] Although the present application is disclosed with the preferred embodiments as above, it is not limited to the present application, and any minor modification, equivalent change and modification made by any person skilled in the art according to the above embodiments of the present application are within the protection scope of the technical solution of the present application.​​​​

Claims

1. A high-precision image sub-region color display method, characterized in that, The method integrates preset window levels and window widths, integrates the corresponding transformation curves of all preset window levels and window widths, transforms the input image data through the transformation function, and displays the transformed image in different colors according to the preset window levels and window widths, so that the image pixels are completely displayed with distinction. The transformation function formula is: wherein, is the first group preset window level of the window level; is the first group window width data of the window width value, the order value of the preset window level and the window width data is corresponding to the color of the pixel; , and satisfies ; is the length of the pixel data in the image supported by the display device; The formula indicates that if the input image pixel gray scale is , the monochrome with the brightness value of is output, and the specific color is determined by the preset window position and window width sequence. Then, in the observer's attention area, the brightness value of the pixel at the attention point in the observer's attention area is adjusted by using the in-area follow-up coloring method, so that the visual system resolution of the observer is maximized, and the image details are better distinguished. Meanwhile, in the observer's attention area, the in-area step difference adjustment technology is used to adjust the step difference between the gray values of the detail pixels in the attention area, so that the observer can effectively distinguish the image details or correctly identify the image contour. The in-area follow-up coloring method is realized by moving the coloring corresponding transformation curve left and right along the horizontal axis. The specific method is: after the transformation formula is transformed and displayed, an observer selects a single color region, if the coordinate of the observer's focus point is , the luminance value of the color of the image pixel gray value after transformation is ; if the luminance value is relatively bright or dark for the observer, the luminance value of the pixel of a certain point in the single color region is taken as the middle value of the luminance value that the image display can present ; If at this time then let , ; Then the Substitute the transformation formula, the brightness of all the pixel points in the area where the observer's focus is located is transformed, while the brightness of the pixel points in other areas does not change. If , let , then into the transformation formula, the luminance of all the pixel points in the area where the observer's focus is located will be transformed, while the luminance of the pixel points in other areas will not change. When the attention point of the observer changes again, the process of the in-area follow-up coloring is repeated by moving the coloring corresponding transformation curve left and right along the horizontal axis.

2. The method of claim 1, wherein the method is characterized by: In the transformation curve, if the transformation curves corresponding to adjacent preset window levels and window widths overlap, the image data with overlapping gray values is uniformly processed according to the rule of left curve or right curve transformation; when the observer observes, the principle of attention area priority is adopted, that is, if the transformation curves of the area where the attention point of the observer is located and the adjacent area overlap, the gray values of the overlapping image data are transformed according to the transformation curve of the area where the attention point is located.

3. The method of claim 1, wherein the method further comprises: determining a color of each pixel in the image based on the color of the pixel in the reference image and the color of the pixel in the target image. The observer can adjust the coloring scheme of each area by himself / herself, but should follow the principle that different areas corresponding to different transformation curves should be colored in different colors, and different areas corresponding to the same transformation curve should be colored in the same color.

4. The method of claim 1, wherein the method is characterized by: In the in-area follow-up coloring method, the blind area and the overlapping area caused by the left and right movement of the in-area data corresponding transformation curve are processed according to the principle of attention area priority for the overlapping area, and are processed by coloring with a fixed color for the blind area, or can be displayed through a pre-set auxiliary transformation curve, which is formed by cyclically moving the original transformation curve.

5. The method of claim 1, wherein the method is characterized by: The in-area step difference adjustment technology takes the brightness of the pixel at the attention point in the current attention area as the window level, and then adjusts the window width value corresponding to the transformation curve corresponding to the color of the area.

Images