Nipple recognition method and system based on pressure sensing array

Abstract

The present application relates to a nipple recognition method and system based on pressure sensing array, comprising the following steps: pressing a pressure sensing array on the front of a breast, the pressure sensing array outputs a two-dimensional array of hardness data, segmenting a breast image according to the two-dimensional array to obtain multiple nipple candidate regions; preprocessing each nipple candidate region, respectively performing an inflation operation and an erosion operation on the preprocessed nipple candidate region to obtain multiple regions HB and multiple regions HC; calculating a nipple confidence index according to the pressure data of regions HB and HC, wherein the region with the largest nipple confidence index is recognized as a nipple. The present application aims to automatically and accurately identify the position of a nipple.

Description

Technical Field

[0001] This invention relates to the field of nipple recognition technology, and in particular to a nipple recognition method and system based on a pressure sensor array. Background Technology

[0002] When examining breast tissue, it is usually divided into five regions. Based on the nipple as the center, a horizontal line and a vertical line are added to divide the breast tissue into four regions: upper outer, lower outer, upper inner, and lower inner. The nipple and areola are then added as the central region.

[0003] Since breast diseases occur in areas centered on the nipple, accurate nipple location is of significant clinical importance for breast tissue examination. However, current methods of manually marking the nipple rely on the physician's personal experience and are very time-consuming. Therefore, an automated and accurate method for determining nipple location is needed. Summary of the Invention

[0004] The purpose of this invention is to automatically and accurately identify nipple location, and to provide a nipple identification method and system based on a pressure sensor array.

[0005] To achieve the above-mentioned objectives, the embodiments of the present invention provide the following technical solutions:

[0006] The nipple identification method based on pressure sensor array includes the following steps:

[0007] Step 1: Press the pressure sensor array onto the front of the breast. The pressure sensor array outputs a two-dimensional array of hardness data. The breast image is segmented according to the two-dimensional array to obtain multiple nipple candidate areas.

[0008] Step 2: Pre-process each nipple candidate area, and perform expansion and corrosion operations on the pre-processed nipple candidate areas to obtain multiple areas HB and multiple areas HC respectively;

[0009] Step 3: Calculate the nipple confidence index based on the pressure data of region HB and region HC. The region with the highest nipple confidence index is identified as the nipple.

[0010] A nipple recognition system based on a pressure sensor array, applied to the recognition system described above, includes:

[0011] A pressure sensor array is used to output hardness data after being pressed on the front of the breast;

[0012] The image segmentation module is used to segment the breast image based on a two-dimensional array of hardness data output by the pressure sensor array to obtain multiple nipple candidate regions.

[0013] The expansion and corrosion processing module is used to pre-process each nipple candidate area, and then perform expansion and corrosion operations on the pre-processed nipple candidate areas to obtain multiple areas HB and multiple areas HC.

[0014] The confidence index calculation module is used to calculate the nipple confidence index based on the pressure data of region HB and region HC, where the region with the highest nipple confidence index is identified as the nipple.

[0015] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0016] This invention determines the candidate nipple region using hardness data. When segmenting the breast image, the image edges are smoothed to facilitate more accurate subsequent expansion and erosion operations. Furthermore, separate structuring elements are set for expansion and erosion operations to distinguish the nipple from other hard objects as much as possible. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart of the method of the present invention;

[0019] Figure 2 This is a block diagram of the system modules of the present invention. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0021] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this invention, the terms "first," "second," etc., are used only for distinguishing descriptions and should not be construed as indicating or implying relative importance, or suggesting any such actual relationship or order between these entities or operations. Additionally, the terms "connected," "linked," etc., can refer to a direct connection between elements or an indirect connection via other elements.

[0022] Example 1:

[0023] This invention is achieved through the following technical solutions, such as... Figure 1 As shown, the nipple recognition method based on a pressure sensor array includes the following steps:

[0024] Step 1: Press the pressure sensor array onto the front of the breast. The pressure sensor array outputs a two-dimensional array of hardness data. The breast image is segmented according to the two-dimensional array to obtain multiple nipple candidate areas.

[0025] Assuming the pressure sensor array outputs k hardness data points, the images containing hardness data points greater than a first threshold or less than a second threshold are segmented to obtain n breast images, where n ≤ k, and the first threshold < the second threshold. Setting a first threshold means that images with hardness data less than the first threshold cannot be nipples; setting a second threshold means that images with hardness data greater than the second threshold cannot be nipples. However, during breast tissue examination, in addition to the nipple corresponding to the hardness data, the pressure sensor may also detect breast nodules, hyperplasia, breast scars, and other hard objects. The hardness data of these objects differs from that of the nipple, and they will also appear differently in the images.

[0026] Each breast image (img_data) is converted to grayscale to obtain the breast image (gray_data); a horizontal two-dimensional convolution C is then applied. x Vertical two-dimensional convolution C y Edge processing is performed on the breast image gray_data to obtain a 3×3 pixel matrix centered at the edge pixels.

[0027] The horizontal two-dimensional convolution C x for:

[0028]

[0029] The vertical two-dimensional convolution C y for:

[0030]

[0031] The 3×3 pixel matrix E centered on the edge pixels is:

[0032]

[0033] Where E i,j This represents the pixel value of the edge pixel.

[0034] Calculate the minimum pixel value in each row of the pixel matrix E:

[0035]

[0036] Among them, row i-1 row i row i+1 These are the minimum pixel values ​​for the (i-1)th row, the ith row, and the (i+1)th row, respectively.

[0037] Calculate the minimum pixel value row in pixel matrix E:

[0038] row = min(row) i-1 min(row) i ,row i+1 ))

[0039] Use the point with the smallest pixel value row in the pixel matrix E as the new edge pixel point f(x,y).

[0040] Perform edge smoothing on the smallest pixel f(x,y):

[0041]

[0042] Among them, G f(x,y) G represents the pixel value of the edge pixel after edge smoothing. img_data For the pixel values ​​of the breast image img_data, G gray_data The pixel values ​​of the gray_data image of the breast;

[0043] After smoothing all edge pixels, adjacent edge pixels form a nipple candidate region D. f = {f(x,y)|x∈n1,y∈n2}, where n1 and n2 are the number of pixel rows and columns of the candidate nipple region, respectively.

[0044] The edges of images segmented by conventional methods are jagged. In order to make subsequent dilation and erosion operations on the candidate nipple region more accurate, this step performs further edge smoothing on the segmented image.

[0045] Step 2: Pre-process each nipple candidate area, and perform expansion and corrosion operations on the pre-processed nipple candidate areas to obtain multiple areas HB and multiple areas HC.

[0046] The purpose of the dilation operation is to find the local maximum, that is, to calculate the maximum value of the pixels in the area covered by the structuring element in the image, and to assign this maximum value to the pixel specified at the center of the current structuring element.

[0047] Let b(s,t) be the first structuring element, and use the first structuring element D. b ={b(s,t)|s∈m,t∈m} for the nipple candidate region D f Perform a dilation operation, where m is the number of pixel rows and columns in the first structuring element:

[0048]

[0049] in This represents the region HB after the expansion operation.

[0050] The erosion operation is the dual operation of the dilation operation. It calculates the minimum pixel value of the region covered by the structuring element in the image and sets the pixel at the center of the current structuring element to the minimum value.

[0051] Let c(u,v) be the second structuring element, and use the second structuring element D. c ={c(u,v)|u∈m`,t∈m`} for the nipple candidate region D f Perform the erosion operation, where m` represents the number of pixel rows and columns of the second structuring element:

[0052]

[0053] in The region HC represents the area after the corrosion operation.

[0054] m and m' can be equal or unequal. Since the connection regions between each hard object and the nipple may be very close, the first structural element D is used to distinguish these connection regions. f With the second structural element D c Setting it separately can better highlight the connected areas, allowing for precise differentiation of the areas where various hard objects or nipples are located.

[0055] Step 3: Calculate the nipple confidence index based on the pressure data of region HB and region HC. The region with the highest nipple confidence index is identified as the nipple.

[0056] The pressure data of regions HB and HC corresponding to the same nipple candidate area were calculated to obtain the average pressure p of region HB. HB The average pressure p in region HCHC ;

[0057] Calculate the nipple confidence index Z for the candidate nipple region:

[0058]

[0059] The candidate nipple region with the largest nipple confidence index Z was selected as the nipple to be identified.

[0060] Please see Figure 2 The present invention also proposes a nipple recognition system based on a pressure sensor array, applied to the above-mentioned recognition method, comprising:

[0061] A pressure sensor array is used to output hardness data after being pressed on the front of the breast;

[0062] The image segmentation module is used to segment the breast image based on a two-dimensional array of hardness data output by the pressure sensor array to obtain multiple nipple candidate regions.

[0063] The expansion and corrosion processing module is used to pre-process each nipple candidate area, and then perform expansion and corrosion operations on the pre-processed nipple candidate areas to obtain multiple areas HB and multiple areas HC.

[0064] The confidence index calculation module is used to calculate the nipple confidence index based on the pressure data of region HB and region HC, where the region with the highest nipple confidence index is identified as the nipple.

[0065] The image segmentation module includes a pixel update unit and an edge correction unit, wherein...

[0066] The pixel update unit is used to segment the image containing hardness data that is greater than a first threshold or less than a second threshold based on the output of k hardness data from the pressure sensor array, and obtain n breast images, where n≤k and the first threshold<the second threshold.

[0067] Each breast image (img_data) is converted to grayscale to obtain the breast image (gray_data); a horizontal two-dimensional convolution C is then applied. x Vertical two-dimensional convolution C y Edge processing is performed on the breast image gray_data to obtain a 3×3 pixel matrix centered at the edge pixels;

[0068] The horizontal two-dimensional convolution C x for:

[0069]

[0070] The vertical two-dimensional convolution C y for:

[0071]

[0072] The 3×3 pixel matrix E centered on the edge pixels is:

[0073]

[0074] Where E i,j Represents the pixel value of the edge pixel;

[0075] Calculate the minimum pixel value in each row of the pixel matrix E:

[0076]

[0077] Among them, row i-1 row i row i+1 These are the minimum pixel values ​​for the (i-1)th row, the ith row, and the (i+1)th row, respectively.

[0078] Calculate the minimum pixel value row in pixel matrix E:

[0079] row = min(row) i-1 min(row) i ,row i+1 ))

[0080] Use the point with the smallest pixel value row in the pixel matrix E as the new edge pixel point f(x,y);

[0081] The edge correction unit is used to perform edge smoothing processing on the smallest pixel point f(x,y):

[0082]

[0083] Among them, G f(x,y) G represents the pixel value of the edge pixel after edge smoothing. img_data For the pixel values ​​of the breast image img_data, G gray_data The pixel values ​​of the gray_data image of the breast;

[0084] After smoothing all edge pixels, adjacent edge pixels form a nipple candidate region D. f = {f(x,y)|x∈n1,y∈n2}, where n1 and n2 are the number of pixel rows and columns of the candidate nipple region, respectively.

[0085] The expansion corrosion treatment module includes an expansion treatment unit and a corrosion treatment unit, wherein...

[0086] The expansion processing unit is used to set b(s,t) as the first structural element and use the first structural element D. b ={b(s,t)|s∈m,t∈m} for the nipple candidate region D f Perform a dilation operation, where m is the number of pixel rows and columns in the first structuring element:

[0087]

[0088] in This represents the region HB after the expansion operation;

[0089] The corrosion processing unit is used to set c(u,v) as the second structural element and use the second structural element D. c ={c(u,v)|u∈m`,t∈m`} for the nipple candidate region D f Perform the erosion operation, where m` represents the number of pixel rows and columns of the second structuring element:

[0090]

[0091] in The region HC represents the area after the corrosion operation.

[0092] The confidence index calculation module calculates the pressure data of region HB and region HC corresponding to the same nipple candidate region, respectively, to obtain the average pressure p of region HB. HB The average pressure p in region HC HC ;

[0093] Calculate the nipple confidence index Z for the candidate nipple region:

[0094]

[0095] The candidate nipple region with the largest nipple confidence index Z was selected as the nipple to be identified.

[0096] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A nipple recognition method based on a pressure sensor array, characterized in that: Includes the following steps: Step 1: Press the pressure sensor array onto the front of the breast. The pressure sensor array outputs a two-dimensional array of hardness data. The breast image is segmented according to the two-dimensional array to obtain multiple nipple candidate areas. Step 2: Pre-process each nipple candidate area, and perform expansion and corrosion operations on the pre-processed nipple candidate areas to obtain multiple areas HB and multiple areas HC respectively; Step 3: Calculate the nipple confidence index based on the pressure data of region HB and region HC. The region with the highest nipple confidence index is identified as the nipple. The step of calculating the nipple confidence index based on the pressure data of region HB and region HC, wherein the region with the highest nipple confidence index is identified as the nipple, includes: The pressure data of regions HB and HC corresponding to the same nipple candidate area were calculated to obtain the average pressure p of region HB. HB The average pressure p in region HC HC ; Calculate the nipple confidence index Z for the candidate nipple region: The candidate nipple region with the largest nipple confidence index Z was selected as the nipple to be identified.

2. The nipple recognition method based on a pressure sensor array according to claim 1, characterized in that: The step of segmenting a breast image according to a two-dimensional array of hardness data output by the pressure sensing array to obtain multiple candidate nipple regions includes: The pressure sensor array outputs k hardness data. The images containing hardness data that are greater than the first threshold and less than the second threshold are segmented to obtain n breast images, where n ≤ k and the first threshold < the second threshold. Each breast image (img_data) is converted to grayscale to obtain the breast image (gray_data); a horizontal two-dimensional convolution C is then applied. x Vertical two-dimensional convolution C y Edge processing is performed on the breast image gray_data to obtain a 3×3 pixel matrix centered at the edge pixels; The horizontal two-dimensional convolution C x for: The vertical two-dimensional convolution C y for: The 3×3 pixel matrix E centered on the edge pixels is: Where E i,j Represents the pixel value of the edge pixel; Calculate the minimum pixel value in each row of the pixel matrix E: Among them, row i-1 row i row i+1 These are the minimum pixel values ​​for the (i-1)th row, the ith row, and the (i+1)th row, respectively. Calculate the minimum pixel value row in pixel matrix E: Use the point with the smallest pixel value row in the pixel matrix E as the new edge pixel point f(x,y), and perform edge smoothing on the smallest pixel point f(x,y): Among them, G f(x,y) G represents the pixel value of the edge pixel after edge smoothing. img_data For the pixel values ​​of the breast image img_data, G gray_data The pixel values ​​of the breast image gray_data; After smoothing all edge pixels, adjacent edge pixels form a nipple candidate region D. f ={f(x,y)|x∈n1,y∈n2}, where n1 and n2 are the number of pixel rows and columns of the candidate nipple region, respectively.

3. The nipple recognition method based on a pressure sensor array according to claim 2, characterized in that: The steps of pre-processing each candidate nipple region, and then performing expansion and etching operations on the pre-processed candidate nipple regions to obtain multiple regions HB and multiple regions HC, include: Let b(s,t) be the first structuring element, and use the first structuring element D. b ={b(s,t)|s∈m,t∈m} represents the candidate nipple region D. f Perform a dilation operation, where m is the number of pixel rows and columns in the first structuring element: in This represents the region HB after the expansion operation; Let c(u,v) be the second structuring element, and use the second structuring element D. c ={c(u,v)|u∈m`,t∈m`} represents the candidate nipple region D. f Perform the erosion operation, where m` represents the number of pixel rows and columns of the second structuring element: in The region HC represents the area after the corrosion operation.

4. A nipple recognition system based on a pressure sensor array, characterized in that: The identification method applied to any one of claims 1-3 includes: A pressure sensor array is used to output hardness data after being pressed on the front of the breast; The image segmentation module is used to segment the breast image based on a two-dimensional array of hardness data output by the pressure sensor array to obtain multiple nipple candidate regions. The expansion and corrosion processing module is used to pre-process each nipple candidate area, and then perform expansion and corrosion operations on the pre-processed nipple candidate areas to obtain multiple areas HB and multiple areas HC. The confidence index calculation module is used to calculate the nipple confidence index based on the pressure data of region HB and region HC, where the region with the largest nipple confidence index is identified as the nipple; The confidence index calculation module calculates the pressure data of region HB and region HC corresponding to the same nipple candidate region, respectively, to obtain the average pressure p of region HB. HB The average pressure p in region HC HC ; Calculate the nipple confidence index Z for the candidate nipple region: The candidate nipple region with the largest nipple confidence index Z was selected as the nipple to be identified.

5. The nipple recognition system based on a pressure sensor array according to claim 4, characterized in that: The image segmentation module includes a pixel update unit and an edge correction unit, wherein... The pixel update unit is used to segment the image containing hardness data that is greater than a first threshold and less than a second threshold based on the output of k hardness data from the pressure sensor array, and obtain n breast images, where n≤k and the first threshold<the second threshold. Each breast image (img_data) is converted to grayscale to obtain the breast image (gray_data); a horizontal two-dimensional convolution C is then applied. x Vertical two-dimensional convolution C y Edge processing is performed on the breast image gray_data to obtain a 3×3 pixel matrix centered at the edge pixels; The horizontal two-dimensional convolution C x for: The vertical two-dimensional convolution C y for: The 3×3 pixel matrix E centered on the edge pixels is: Where E i,j Represents the pixel value of the edge pixel; Calculate the minimum pixel value in each row of the pixel matrix E: Among them, row i-1 row i row i+1 These are the minimum pixel values ​​for the (i-1)th row, the ith row, and the (i+1)th row, respectively. Calculate the minimum pixel value row in pixel matrix E: Use the point with the smallest pixel value row in the pixel matrix E as the new edge pixel point f(x,y); The edge correction unit is used to perform edge smoothing processing on the smallest pixel point f(x,y): Among them, G f(x,y) G represents the pixel value of the edge pixel after edge smoothing. img_data For the pixel values ​​of the breast image img_data, G gray_data The pixel values ​​of the breast image gray_data; After smoothing all edge pixels, adjacent edge pixels form a nipple candidate region D. f ={f(x,y)|x∈n1,y∈n2}, where n1 and n2 are the number of pixel rows and columns of the candidate nipple region, respectively.

6. The nipple recognition system based on a pressure sensor array according to claim 5, characterized in that: The expansion corrosion treatment module includes an expansion treatment unit and a corrosion treatment unit, wherein... The expansion processing unit is used to set b(s,t) as the first structural element and use the first structural element D. b ={b(s,t)|s∈m,t∈m} represents the candidate nipple region D. f Perform a dilation operation, where m is the number of pixel rows and columns in the first structuring element: in This represents the region HB after the expansion operation; The corrosion processing unit is used to set c(u,v) as the second structural element and use the second structural element D. c ={c(u,v)|u∈m`,t∈m`} represents the candidate nipple region D. f Perform the erosion operation, where m` represents the number of pixel rows and columns of the second structuring element: in The region HC represents the area after the corrosion operation.

Images