Image processing method and system in electronic colposcope

Abstract

The image processing method in the electronic colposcope provided by the present invention first divides the image to be processed into three parts by using an image segmentation method, and uses a high-precision interpolation method for the key observation area, and uses a bilinear interpolation method with a fast calculation speed for interpolation in the non-key observation area, that is, the first interpolation method is used to interpolate the first part, the second interpolation method is used to interpolate the second part, and the bilinear interpolation method is used to interpolate the third part. Since interpolation easily causes jagged edges in the boundary area, the inserted pixels of the first boundary area and the second boundary area in the intermediate image obtained after the above interpolation are corrected. The method provided by the present invention not only improves the image processing speed of the electronic colposcope, but also can obtain high-quality images.

Description

Technical Field

[0001] The present application relates to the technical field of image processing, and in particular to image processing in an electronic colposcope. Background Art

[0002] Electronic colposcope (Colposcopy) is an important device for examining the female vagina and cervix. Doctors can use electronic colposcope to check whether there are any abnormalities in the vagina and cervix. The basic principle of electronic colposcope is to insert the charge coupled device (CCD) and LED cold light source deep into the patient's vagina, transmit the image data collected by the CCD to the computer, and then display the image inside the vagina on the monitor through computer processing. Electronic colposcope can magnify the image dozens of times. Through electronic colposcope, tiny lesions, erosions, etc. can be displayed on the monitor. Because it is a non-invasive examination method, electronic colposcope has been widely used in the examination of gynecological diseases.

[0003] When using an electronic colposcope for gynecological examination, images need to be processed in real time, and image processing is the most resource-intensive operation of the electronic colposcope system, especially for some processors with low performance. If a complex image processing method is used, the image quality will be improved, but the processing speed of the processor cannot match that of the CCD, which may cause jamming; if a simple image processing method is used, the processing speed will be improved, but the image quality will be reduced. For example, when the doctor enlarges the image, obvious jagged edges will appear.

[0004] Image quality and processing speed are contradictory and cannot be satisfied at the same time. Currently, most of the focus is on improving image processing methods, but so far there is no simple, fast method that can meet the requirements of high image quality. How to process images quickly with limited resources and obtain images that meet the requirements of doctors is an urgent problem that needs to be solved. Summary of the invention

[0005] The present invention aims to solve the problem of obtaining an image that meets the doctor's requirements under limited resource conditions with an electronic colposcope. The present invention improves the interpolation of images in the electronic colposcope and increases the image interpolation speed in the electronic colposcope system without affecting the quality of the image that the doctor is interested in.

[0006] In one aspect, the present invention provides an image processing method in an electronic colposcope, wherein the electronic colposcope adopts a full-frame CCD image sensor, and the method comprises the following steps:

[0007] S1, an image processing unit obtains an image to be processed, and uses an image segmentation method to segment the image to be processed into three parts, wherein the first part is the cervix, the second part is the cervix, and the third part is the vaginal wall, and obtains a first boundary line and a second boundary line of the image to be processed, wherein the first boundary line is the boundary between the vaginal wall and the cervix, and the second boundary line is the boundary between the cervix and the cervix, a first boundary area is set with the first boundary line as the center, and the width of the first boundary area is m pixels, and a second boundary area is set with the second boundary line as the center, and the width of the second boundary area is n pixels, where m and n are positive integers;

[0008] S2, interpolating the first part using a first interpolation method, interpolating the second part using a second interpolation method, and interpolating the third part using a bilinear interpolation method, and splicing the interpolated images into an intermediate image;

[0009] S3, correcting the interpolated pixels in the first boundary area and the second boundary area in the intermediate image.

[0010] Preferably, the correcting the inserted pixels of the first boundary area and the second boundary area in the intermediate image specifically comprises the following steps:

[0011] S31, taking the center position of the first part as the center of the circle, interpolating the boundary area along the direction from the outside to the center of the circle;

[0012] S32, compare the pixel point interpolation value t1 of S31 with the pixel point interpolation value t2 corresponding to the intermediate image spliced ​​in S2, if the difference between the two is less than the threshold, then use the average value of t1 and t2 as the interpolation value of the pixel point; otherwise, execute S33;

[0013] S33, interpolate the boundary area along the direction of the boundary line to obtain the pixel interpolation value t3. If |t1-t3|>|t2-t3|, then the average value of t2 and t3 is used as the interpolation value of the pixel point; otherwise, the average value of t1 and t3 is used as the interpolation value of the pixel point.

[0014] Preferably, the interpolation of the boundary region along the direction from the outside to the center of the circle is specifically: taking the direction from the outside to the center of the circle as the gradient direction, and interpolating the boundary region using a gradient interpolation method;

[0015] The interpolation of the boundary along the direction of the boundary line specifically includes: taking the boundary line as the gradient direction and interpolating the boundary area by using the gradient interpolation method.

[0016] Preferably, the first boundary area and the second boundary area are annular.

[0017] Preferably, the first interpolation method and / or the second interpolation method is a color ratio constant method.

[0018] In addition, the present invention also provides an image processing system in an electronic colposcope, the system is applied to image processing of a colposcope using a full-frame CCD image sensor, and the system includes the following modules:

[0019] An image segmentation module, wherein the image processing unit obtains the image to be processed, and uses an image segmentation method to divide the image to be processed into three parts, wherein the first part is the uterine os, the second part is the cervix, and the third part is the vaginal wall, and obtains a first boundary line and a second boundary line of the image to be processed, wherein the first boundary line is the boundary between the vaginal wall and the cervix, and the second boundary line is the boundary between the cervix and the uterine os, a first boundary area is set with the first boundary line as the center, and the width of the first boundary area is m pixels, and a second boundary area is set with the second boundary line as the center, and the width of the second boundary area is n pixels, where m and n are positive integers;

[0020] An image interpolation module, used to interpolate the first part using a first interpolation method, interpolate the second part using a second interpolation method, interpolate the third part using a bilinear interpolation method, and splice the interpolated images into an intermediate image;

[0021] The boundary correction module is used to correct the inserted pixels of the first boundary area and the second boundary area in the intermediate image.

[0022] Preferably, the correcting the inserted pixels of the first boundary area and the second boundary area in the intermediate image is specifically:

[0023] Taking the center position of the first part as the center of the circle, interpolate the boundary area along the direction from the outside to the center of the circle; compare the interpolation value t1 with the pixel point interpolation value t2 corresponding to the stitched intermediate image, if the difference between the two is less than the threshold, then use the average value of t1 and t2 as the interpolation value of the pixel point; otherwise, interpolate the boundary along the direction of the boundary to obtain the pixel point interpolation value t3, if |t1-t3|>|t2-t3|, then use the average value of t2 and t3 as the interpolation value of the pixel point, otherwise use the average value of t1 and t3 as the interpolation value of the pixel point.

[0024] Preferably, the interpolation of the boundary area along the direction from the outside to the center of the circle is specifically: taking the direction from the outside to the center of the circle as the gradient direction, and interpolating the boundary using a gradient interpolation method;

[0025] The interpolation of the boundary area along the direction of the boundary is specifically: taking the boundary line as the gradient direction and using the gradient interpolation method to interpolate the boundary area.

[0026] Preferably, the first boundary area and the second boundary area are annular, and the first interpolation method and / or the second interpolation method is a color ratio constant method.

[0027] Finally, the present invention also provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method described above is implemented.

[0028] The electronic colposcopy image processing method of the present invention adopts different interpolation methods for different areas. For areas that need to be focused on, an interpolation method that can improve the resolution is adopted. For less important areas, a simple bilinear interpolation method is adopted, thereby avoiding the existing interpolation method from interpolating the entire image and improving the processing speed of image interpolation. At the same time, when interpolating, the edge area is prone to jagged edges. In this regard, the present invention can also correct the boundaries in the intermediate image to avoid boundary blur. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0030] Figure 1 A structural diagram of an electronic colposcope system provided by an embodiment of the present invention;

[0031] Figure 2 A typical example image taken for a digital colposcope;

[0032] Figure 3 An image processing flow chart provided by an embodiment of the present invention;

[0033] Figure 4 A flow chart of boundary correction provided by an embodiment of the present invention;

[0034] Figure 5 A schematic diagram of boundary correction provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0035] In this article, relational terms such as first and second, etc. are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "comprise", "include" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the statement "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0036] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0037] Figure 1 An electronic colposcope system is shown, which mainly includes a CCD image sensor, an LED cold light source, a processor, a memory, and a display, wherein an image processing program runs in the processor. The system also includes, for example, a power supply and a video decoding unit between the CCD and the processor, which are not shown.

[0038] like Figure 2 As shown in the figure, it is an image taken by an electronic colposcope. Since the electronic colposcope is mainly used for the examination of the cervix, the image can be divided into three parts, namely the vaginal wall, the cervix, and the cervix. The cervix and the cervix are the focus of the examination. In order to find subtle lesions, the cervix and the cervix need to be enlarged, so that it is convenient for doctors to observe the conditions of the cervix and the cervix. Most of the existing electronic colposcope systems use embedded chips as processing elements for image processing, such as FPGA, DSP, ARM, etc. Since the images taken by CCD are high-definition, the resources of these processing elements are limited, and they also need to respond to the doctor's operations such as freezing the image. In actual applications, the image processing is not timely.

[0039] like Figure 3 As shown, the present invention first provides an image processing method in an electronic colposcope, wherein the electronic colposcope adopts a full-frame CCD image sensor, and the method comprises the following steps:

[0040] S1, an image processing unit obtains an image to be processed, and uses an image segmentation method to segment the image to be processed into three parts, wherein the first part is the cervix, the second part is the cervix, and the third part is the vaginal wall, and obtains a first boundary line and a second boundary line of the image to be processed, wherein the first boundary line is the boundary between the vaginal wall and the cervix, and the second boundary line is the boundary between the cervix and the cervix, a first boundary area is set with the first boundary line as the center, and the width of the first boundary area is m pixels, and a second boundary area is set with the second boundary line as the center, and the width of the second boundary area is n pixels, where m and n are positive integers;

[0041] S2, interpolating the first part using a first interpolation method, interpolating the second part using a second interpolation method, and interpolating the third part using a bilinear interpolation method, and splicing the interpolated images into an intermediate image;

[0042] S3, correcting the interpolated pixels in the first boundary area and the second boundary area in the intermediate image.

[0043] like Figure 4 As shown, in a specific embodiment, the correction of the interpolated pixels of the first boundary area and the second boundary area in the intermediate image specifically includes the following steps:

[0044] S31, taking the center position of the first part as the center of the circle, interpolating the boundary area along the direction from the outside to the center of the circle;

[0045] S32, compare the pixel point interpolation value t1 of S31 with the pixel point interpolation value t2 corresponding to the intermediate image spliced ​​in S2, if the difference between the two is less than the threshold, then use the average value of t1 and t2 as the interpolation value of the pixel point; otherwise, execute S33;

[0046] S33, interpolate the boundary area along the direction of the boundary line to obtain the pixel interpolation value t3. If |t1-t3|>|t2-t3|, then the average value of t2 and t3 is used as the interpolation value of the pixel point; otherwise, the average value of t1 and t3 is used as the interpolation value of the pixel point.

[0047] like Figure 5 As shown in the figure, the boundary area is between the two curves on the left, and the dotted line is the boundary line. The dark dots represent the original pixels, and the white dots represent the pixels that need to be interpolated. Figure 5 The dashed line shown is the first boundary line, the left side of the dashed line is the cervix, and the right side of the dashed line is the vaginal wall. In the middle image, the right side of the dashed line belongs to the third part, Figure 5 The middle pixel 1 belongs to the third part. The white point on the right side of the dotted line is interpolated using the bilinear interpolation method, and the interpolation value t2 is obtained. Similarly, the pixel value on the left side of the dotted line can be obtained.

[0048] The pixel point 1 is interpolated along the direction from the center of the circle using the gradient interpolation method to obtain the interpolation value t1. The difference between t1 and t2 is compared. If the difference is large, it indicates that the position point may be a point in the boundary area where a pixel may change significantly. In this regard, the pixel point 1 is further interpolated according to the first boundary, that is, the direction of the dotted line, to obtain the interpolation value t3. If |t1-t3|>|t2-t3|, the average value of t2 and t3 is used as the interpolation value of the pixel point, otherwise the average value of t1 and t3 is used as the interpolation value of the pixel point. Then the interpolation value of the boundary area is corrected. This method can effectively avoid the appearance of aliasing.

[0049] In a specific embodiment, the interpolation of the boundary region along the direction from the outside to the center of the circle is specifically: taking the direction from the outside to the center of the circle as the gradient direction, and interpolating the boundary region using a gradient interpolation method;

[0050] The interpolation along the direction of the boundary is specifically performed by taking the boundary line as the gradient direction and interpolating the boundary area by using the gradient interpolation method.

[0051] In a specific embodiment, the first boundary area and the second boundary area are annular.

[0052] The cervix and uterine ostium are prone to lesions and are also key observation areas for doctors. For the first and second parts, an image insertion method that can obtain values ​​closer to the original values ​​is used. For non-key areas, a simple insertion method is used to increase the speed of image processing. There are many image insertion methods, such as nearest neighbor interpolation, linear interpolation, bilinear interpolation, stereo convolution interpolation, gradient-based interpolation, adaptive interpolation, etc. A common point of interpolation methods is that the processing speed is inversely proportional to the imaging quality. In a specific embodiment, the first interpolation method and / or the second interpolation method is a color ratio constant method or a stereo convolution interpolation method.

[0053] In a specific embodiment, before step S3, the method further includes: monitoring the processor of the electronic colposcope system to obtain the occupancy rate of the processor, and if the occupancy rate is greater than a first preset value, taking the intermediate image as the final interpolated image and not executing S3. This can reduce the probability of image browsing freezes. The processor is FPGA, DSP or ARM.

[0054] In another embodiment, before step S3, it further includes: obtaining the change of the processor occupancy rate of the thread executing S3 before and after the thread executes the last image correction, if the processor occupancy rate is greater than the first preset value, and the change is greater than the second preset value, the intermediate image is used as the final interpolated image, and S3 is no longer executed, if the change is less than the second preset value, S3 is continued to be executed. When the processor occupancy rate is large, but executing S3 does not occupy too much system resources, it means that executing or not executing S3 does not have much impact on the execution speed of the system. In this case, S3 can be continued to be executed.

[0055] In another embodiment, before executing S3, it also includes: monitoring the processor of the electronic colposcope system to obtain the occupancy rate of the processor, and if the occupancy rate is greater than a first preset value, no correction is performed on the first boundary area or the second boundary area.

[0056] Secondly, the present invention also provides an image processing system in an electronic colposcope, the system is applied to the image processing of a colposcope using a full-frame CCD image sensor, and the system includes the following modules:

[0057] An image segmentation module, wherein the image processing unit obtains the image to be processed, and uses an image segmentation method to divide the image to be processed into three parts, wherein the first part is the uterine os, the second part is the cervix, and the third part is the vaginal wall, and obtains a first boundary line and a second boundary line of the image to be processed, wherein the first boundary line is the boundary between the vaginal wall and the cervix, and the second boundary line is the boundary between the cervix and the uterine os, a first boundary area is set with the first boundary line as the center, and the width of the first boundary area is m pixels, and a second boundary area is set with the second boundary line as the center, and the width of the second boundary area is n pixels, where m and n are positive integers;

[0058] An image interpolation module, used to interpolate the first part using a first interpolation method, interpolate the second part using a second interpolation method, interpolate the third part using a bilinear interpolation method, and splice the interpolated images into an intermediate image;

[0059] The boundary correction module is used to correct the inserted pixels of the first boundary area and the second boundary area in the intermediate image.

[0060] Preferably, the correcting the inserted pixels of the first boundary area and the second boundary area in the intermediate image is specifically:

[0061] Taking the center position of the first part as the center of the circle, interpolate the boundary area along the direction from the outside to the center of the circle; compare the interpolation value t1 with the pixel point interpolation value t2 corresponding to the stitched intermediate image, if the difference between the two is less than the threshold, then use the average value of t1 and t2 as the interpolation value of the pixel point; otherwise, interpolate the boundary along the direction of the boundary to obtain the pixel point interpolation value t3, if |t1-t3|>|t2-t3|, then use the average value of t2 and t3 as the interpolation value of the pixel point, otherwise use the average value of t1 and t3 as the interpolation value of the pixel point.

[0062] Preferably, the interpolation of the boundary area along the direction from the outside to the center of the circle is specifically: taking the direction from the outside to the center of the circle as the gradient direction, and interpolating the boundary using a gradient interpolation method;

[0063] The interpolation of the boundary area along the direction of the boundary is specifically: taking the boundary line as the gradient direction and using the gradient interpolation method to interpolate the boundary area.

[0064] Preferably, the first boundary area and the second boundary area are annular, and the first interpolation method and / or the second interpolation method is a color ratio constant method.

[0065] Finally, the present invention also provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method described above is implemented.

[0066] The various embodiments described in the present invention can be combined to implement the corresponding technical solutions. The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented directly using hardware, software modules executed by a processor, or a combination of the two. The software module can be placed in a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the technical field.

Claims

1. An image processing method in an electronic colposcope, wherein the electronic colposcope adopts a full-frame CCD image sensor, characterized in that: The method comprises the following steps: S1, an image processing unit obtains an image to be processed, and uses an image segmentation method to segment the image to be processed into three parts, wherein the first part is the cervix, the second part is the cervix, and the third part is the vaginal wall, and obtains a first boundary line and a second boundary line of the image to be processed, wherein the first boundary line is the boundary between the vaginal wall and the cervix, and the second boundary line is the boundary between the cervix and the cervix, a first boundary area is set with the first boundary line as the center, and the width of the first boundary area is m pixels, and a second boundary area is set with the second boundary line as the center, and the width of the second boundary area is n pixels, where m and n are positive integers; S2, interpolating the first part using a first interpolation method, interpolating the second part using a second interpolation method, and interpolating the third part using a bilinear interpolation method, and splicing the interpolated images into an intermediate image; S3, correcting the interpolated pixels in the first boundary area and the second boundary area in the intermediate image; The correcting of the interpolated pixels of the first boundary area and the second boundary area in the intermediate image specifically comprises the following steps: S31, taking the center position of the first part as the center of the circle, interpolating the boundary area along the direction from the outside to the center of the circle; S32, compare the pixel point interpolation value t1 of S31 with the pixel point interpolation value t2 corresponding to the intermediate image spliced ​​in S2, if the difference between the two is less than the threshold, then use the average value of t1 and t2 as the interpolation value of the pixel point; otherwise, execute S33; S33, interpolate the boundary area along the direction of the boundary line to obtain the pixel interpolation value t3, if |t1-t3|>|t2-t3|, then use the average value of t2 and t3 as the interpolation value of the pixel point, otherwise use the average value of t1 and t3 as the interpolation value of the pixel point; The interpolation of the boundary region along the direction from the outside to the center of the circle is specifically: taking the direction from the outside to the center of the circle as the gradient direction, and interpolating the boundary region by using the gradient interpolation method; The interpolation of the boundary area along the direction of the boundary line specifically includes: taking the boundary line as the gradient direction and interpolating the boundary using the gradient interpolation method.

2. The method according to claim 1, characterized in that The first boundary area and the second boundary area are annular.

3. The method according to any one of claims 1 to 2, characterized in that: The first interpolation method and / or the second interpolation method is a color ratio constant method.

4. An image processing system in an electronic colposcope, the system being applied to image processing of a colposcope using a full-frame CCD image sensor, characterized in that: The system includes the following modules: An image segmentation module, wherein the image processing unit obtains the image to be processed, and uses an image segmentation method to divide the image to be processed into three parts, wherein the first part is the uterine os, the second part is the cervix, and the third part is the vaginal wall, and obtains a first boundary line and a second boundary line of the image to be processed, wherein the first boundary line is the boundary between the vaginal wall and the cervix, and the second boundary line is the boundary between the cervix and the uterine os, a first boundary area is set with the first boundary line as the center, and the width of the first boundary area is m pixels, and a second boundary area is set with the second boundary line as the center, and the width of the second boundary area is n pixels, where m and n are positive integers; An image interpolation module, used to interpolate the first part using a first interpolation method, interpolate the second part using a second interpolation method, interpolate the third part using a bilinear interpolation method, and splice the interpolated images into an intermediate image; A boundary correction module, used for correcting the inserted pixels of the first boundary area and the second boundary area in the intermediate image; The correcting of the interpolated pixels of the first boundary area and the second boundary area in the intermediate image is specifically: Taking the center position of the first part as the center of the circle, interpolate the boundary area along the direction from the outside to the center of the circle; compare the interpolation value t1 with the pixel point interpolation value t2 corresponding to the stitched intermediate image, if the difference between the two is less than the threshold, then use the average value of t1 and t2 as the interpolation value of the pixel point; otherwise, interpolate the boundary area along the direction of the boundary line to obtain the pixel point interpolation value t3, if |t1-t3|>|t2-t3|, then use the average value of t2 and t3 as the interpolation value of the pixel point, otherwise use the average value of t1 and t3 as the interpolation value of the pixel point; The interpolation of the boundary region along the direction from the outside to the center of the circle is specifically: taking the direction from the outside to the center of the circle as the gradient direction, and interpolating the boundary region by using the gradient interpolation method; The interpolation along the direction of the boundary is specifically: taking the boundary line as the gradient direction and interpolating the boundary area by using the gradient interpolation method.

5. The system according to claim 4, characterized in that The first boundary area and the second boundary area are annular, and the first interpolation method and / or the second interpolation method is a constant color ratio method.

6. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 3 is implemented.

Images