CN116797778A - Region of interest acquisition method, electronic device and storage medium - Google Patents

Abstract

The invention provides a method for acquiring a region of interest, which comprises the following steps: identifying a human body contour line in the target infrared image to obtain a corresponding human body contour line pixel point set PL; identifying a region of interest in the limb and first and second characteristic acupoints CA1 and CA2 in the torso, respectively; acquiring remaining acupoints based on the identified CA1 and CA 2; the trunk is divided into a plurality of interested areas based on the obtained n set acupoints, and the boundaries of the areas positioned on two sides are human body contour lines. The invention also provides electronic equipment and a storage medium. The invention can make the acquired interested region more accurate.

Description

Region of interest acquisition method, electronic device and storage medium

Technical field

The present invention relates to the field of computer application technology, and in particular to a method for obtaining a region of interest, an electronic device and a storage medium.

Background technique

The theory of yin and yang balance of human body meridians and acupuncture points is one of the core theories of traditional Chinese medicine. Infrared thermal imaging of the human body surface can clearly reflect the temperature of each acupoint on the human body and the thermal order of each meridians. Through analysis, the body's constitution and syndrome characteristics can be obtained. For example, according to the theory of traditional Chinese medicine, the human trunk area is divided into multiple set areas, and the functional status of the human body is detected through temperature analysis of these areas. However, the existing area division does not include all areas within the human body contour line. For example, there will be unused areas between the divided areas and the human body contour line, as shown in Figure 1. In this way, there will be a problem of inaccurate analysis results.

Contents of the invention

In view of the above technical problems, the technical solutions adopted by the present invention are:

An embodiment of the present invention provides a method for obtaining a region of interest, which method includes the following steps:

An embodiment of the present invention provides a method for obtaining a region of interest, which method includes the following steps:

S100, identify the human body contour line in the target infrared image, and obtain the corresponding human body contour line pixel point set PL = (PL ₁ , PL ₂ , ..., PL _r , ..., PL _M ), PL _r is the PL th r pixels, r ranges from 1 to M, M is the number of pixels in PL, and the coordinates of PL _r are (x _Lr , y _Lr );

S200, identify m regions of interest in the limbs in the target infrared image and the first characteristic acupuncture point CA1 and the second characteristic acupoint CA2 in the trunk, respectively, and obtain m regions of interest pixel point sets and first characteristic pixels respectively. Point PA1 and the second feature pixel point PA2; among them, the k-th region of interest pixel point set is PROI _k = (PROI _k1 , PROI _k2 ,..., PROI _kg ,..., PROI _kf(k) ), PROI _kg is PROI The g-th pixel in _k , the value of g ranges from 1 to f(k), f(k) PROI The number of pixels in _k , the value of k ranges from 1 to m; CA1 and CA2 are respectively the Among the n set acupuncture points, CA1 and CA2 are located on the central axis of the human body, CA2 is located below CA1, and the pixel coordinates of PA1 and PA2 are (x _A1 , y _A1 ) and (x _A2 , y _A2) respectively. );

S300, based on CA1 and CA2 and the set division rules, divide the composite area surrounded by the contour lines of the head and neck into H areas of interest, where the area of interest located at the outermost side of the composite area includes the corresponding composite area. contour line;

S400, based on the identified CA1 and CA2, obtain the remaining acupoints among the n set acupoints in the target infrared image;

S500: Divide the trunk area surrounded by the contour line of the trunk into F regions of interest based on the acquired n set acupuncture points, wherein the area of interest located at the outermost side of the trunk area includes the corresponding trunk contour line.

An embodiment of the present invention also provides a non-transitory computer-readable storage medium, in which at least one instruction or at least one program is stored, and the at least one instruction or at least one program is loaded and executed by a processor to implement As mentioned above.

An embodiment of the present invention also provides an electronic device, including a processor and the aforementioned non-transitory computer-readable storage medium. The present invention has at least the following beneficial effects:

The method of obtaining the region of interest provided by the embodiment of the present invention can at least extend the region of interest of the human body to the human body contour, thereby obtaining as many regions of interest as possible, thereby making the analysis more accurate.

Description of the drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed to be used in the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can also be obtained based on these drawings without exerting creative efforts.

Figure 1 is a schematic diagram of the existing trunk area of interest division.

Figure 2 is a flow chart of a method for obtaining a region of interest provided by an embodiment of the present invention.

FIG. 3 is a schematic diagram of the division of regions of interest in a composite region provided by an embodiment of the present invention.

Detailed ways

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making creative efforts fall within the scope of protection of the present invention.

An embodiment of the present invention provides a method for obtaining a region of interest, as shown in Figure 2. The method may include the following steps:

S100, identify the human body contour line in the target infrared image, and obtain the corresponding human body contour line pixel point set PL = (PL ₁ , PL ₂ , ..., PL _r , ..., PL _M ), PL _r is the PL th There are r pixels, the value of r is 1 to M, M is the number of pixels in PL, and the coordinates of PL _r are (x _Lr , y _Lr ).

In the embodiment of the present invention, the x-axis of the coordinate system where the target infrared image is located is the transverse direction of the human body, the y-axis is the longitudinal direction of the human body, and the origin of the coordinates can be the lower left corner of the image.

In the embodiment of the present invention, the target infrared image can be obtained from the image library, or the target infrared image can be obtained in real time through an infrared camera device.

Further, in the embodiment of the present invention, PL can be obtained based on the following steps:

S101, identify the human body contour in the target infrared image, and obtain the corresponding human body contour initial pixel point set PLI = (PLI ₁ , PLI ₂ , ..., PLI _r , ..., PLI _M ).

S102, obtain the temperature set corresponding to PLI TCI = (TLI ₁ , TLI ₂ , ..., TLI _r , ..., TLI _M ), where TLI _r is the temperature corresponding to PLI _r .

Those skilled in the art know that any acquisition of the temperature of each pixel point from an infrared image falls within the scope of the present invention.

S103, if TLI _r <T0, execute S104; T0 is the preset temperature value; otherwise, execute S107.

In an exemplary embodiment of the present invention, T0 is a preset average human body temperature threshold, which can be an empirical value, for example, 36.5°. Specifically, T0 can be obtained as follows:

T0=Avg(T1, T2, ..., Tz, ..., TG), Tz is the body temperature of the z-th user, z ranges from 1 to G, and G is the number of users. G users can be obtained randomly, and the body temperature of each user can be measured by a temperature measuring device.

In another illustrative embodiment of the present invention, T0=Avg(t ₁ , t ₂ ,..., _tw ,..., t _N ), where _tw is the area enclosed by the human body outline identified in S101 The temperature corresponding to the w-th pixel, N is the number of pixels in the area enclosed by the human body contour identified in S101. Compared with the previous embodiment, T0 in this embodiment can improve accuracy because the user's body temperature corresponding to the target infrared image is used as the reference temperature.

S104, if _{x LIr} <x ₀ , set x _LIr = x _LIr + b; if x _LIr > x ₀ , set x _LIr = x _LIr -b; The abscissa of , x _LIr is the abscissa of the pixel corresponding to TLI _r , and b is the preset number of pixels; execute S105.

In an exemplary embodiment, b is set to a value that does not affect the accuracy of acupuncture point recognition, for example, b=1.

S105, if the temperature corresponding to the pixel point (x _LIr , y _LIr ) is less than T0, execute S104, that is, continue to adjust the abscissa of the pixel point until the corresponding temperature is greater than T0 or equal to T0; otherwise, execute S106.

S106, set r=r+1, if r≤M, execute S103; otherwise, exit the control program.

The technical effect of S101 to S106 is that due to the different body structure of each user, the human body contour line recognized by the trained image recognition model may be biased. Through S101 to S106, the recognized human body contour line can be adjusted so that Human body contours are more accurate.

S200, identify m regions of interest in the limbs in the target infrared image and the first characteristic acupuncture point CA1 and the second characteristic acupoint CA2 in the trunk, respectively, and obtain m regions of interest pixel point sets and first characteristic pixels respectively. Point PA1 and the second feature pixel point PA2; among them, the k-th region of interest pixel point set is PROI _k = (PROI _k1 , PROI _k2 ,..., PROI _kg ,..., PROI _kf(k) ), PROI _kg is PROI The g-th pixel in _k , the value of g ranges from 1 to f(k), f(k) PROI The number of pixels in _k , the value of k ranges from 1 to m; CA1 and CA2 are respectively the Among the n set acupuncture points, CA1 and CA2 are located on the central axis of the human body, CA2 is located below CA1, and the pixel coordinates of PA1 and PA2 are (x _A1 , y _A1 ) and (x _A2 , y _A2) respectively. ).

In embodiments of the present invention, the human body contours and characteristic acupuncture points in the target infrared image can be identified through the trained image recognition model. Specifically, the trained image recognition model can be obtained by inputting sample images marked with human body contours, regions of interest in limbs, CA1 and CA2 into the neural network model for training. The specific training method can be based on existing technologies. In this embodiment of the present invention, the areas of interest in the limbs may be joint areas such as wrist joints, elbow joints, and knee joints. Due to the symmetry of the limbs, a classification algorithm can be used to identify the area of interest in the limbs. That is, when the area of interest in the limbs is identified, the relationship between the abscissa of the identified area of interest and the abscissa of the characteristic acupoint can be relationship to determine whether it is the left limb or the right limb.

In the case where the n set acupuncture points are acupoints in the frontal trunk area, the target infrared image is a frontal image of the human body, CA1 can be the easily identified Tiantu point, and CA2 can be the easily identified Shenque point. In the case where the n set acupuncture points are acupuncture points in the back torso area, the target infrared image is the back image of the human body, CA1 can be the easily identified Dazhui point, and CA2 can be the easily identified Changqiang point.

In the embodiment of the present invention, the region of interest is a region divided into human body regions based on analysis needs, and can be divided based on analysis needs and traditional Chinese medicine theory.

S300, based on CA1 and CA2 and the set division rules, divide the composite area surrounded by the contour lines of the head and neck into H areas of interest, where the area of interest located at the outermost side of the composite area includes the corresponding composite area. contour line.

S400: Based on the identified CA1 and CA2, obtain the remaining acupoints among the n set acupoints in the target infrared image.

Further, S400 specifically includes:

S402, obtain the two symmetrical pixel coordinates (x ₀₁ , y _A1 ) and (x ₀₂ , y _A1 ) with the ordinate y _A1 from PL, that is, the two intersection points of the first straight line and the human body contour line, the first The straight line is a straight line that passes through the first characteristic acupuncture point and is perpendicular to the central axis of the human body, and _the two symmetrical pixel point coordinates (x ₀₃ , y _A2 ) and (x ₀₄ , y _A2 ) whose ordinate is y A2 are obtained from PL, That is, the two intersection points of the second straight line and the human body contour line. The second straight line is a straight line that passes through the second characteristic acupoint and is perpendicular to the central axis of the human body; among them, x ₀₁ ＜x _A1 ＜x ₀₂ , x ₀₃ ＜x _A2 ＜x ₀₄ .

S404, obtain the pixel coordinates corresponding to (min (x ₀₁ , x ₀₃ ), y _A1 ) or (min (x ₀₂ , x ₀₄ ), y _A1 ) as the coordinates (x _A3 , y _A3 ) of the reference feature point CA3, That is, the coordinates corresponding to the shortest intersection line among the first intersection line and the second intersection line are selected as the coordinates of CA3. The first intersection line is the intersection line between the first straight line and the human body contour line, and the second intersection line is the intersection line of the first straight line and the human body contour line. The intersection between two straight lines and the human body contour.

In the embodiment of the present invention, CA3 may be the shoulder point.

S406, obtain the horizontal length △x＝∣x _A3 -x _A1 ∣/W ₁₃ , and obtain the vertical length △y＝∣y _A2 -y _A1 ∣/V ₁₂ , W ₁₃ is the horizontal length between CA1 and CA3 Quantity, V ₁₂ is the number of inches between CA1 and CA2. W ₁₃ and V ₁₂ can be determined based on traditional Chinese medicine theory.

S408: Using CA1 or CA2 as the reference acupoint, based on Δx and Δy and the position relationship table between the n set acupuncture points, obtain the remaining acupoints among the n set acupuncture points in the target infrared image.

In embodiments of the present invention, the positional relationship between acupoints can be determined based on traditional Chinese medicine theory, such as acupuncture.

Specifically, the n set acupoints may include n1 first acupoints located on the central axis of the human body and n2 second acupoints located on both sides of the central axis of the human body. The positional relationship between the n set acupoints is as follows: It includes a first acupoint location relationship table and a second acupoint location relationship table, where the i-th row of the first acupoint location relationship table includes (C1 _i , h1, P _ci ), and C1 _i is the i-th row among the n1 first acupoints. The ID of the first acupoint, i ranges from 1 to n1, and h1 takes a value of 1 or 2. Among them, h1=1 means that C1 _i is located on the upper side of the reference acupoint, and h1=2 means that C1 _i is located on the upper side of the reference acupoint. On the lower side, P _ci represents the number of straight inches between the i-th first acupoint and the reference acupoint. The j-th row of the second acupoint position relationship table includes (C2 _j , C1 _je , (h1, PX _je ), (h2, PY _je )), and C2 _j is the j-th second acupoint among the n2 second acupoints. ID, C1 _je is the ID of the first acupoint associated with C2 _j , the value of h2 is 3 or 4, h2=3 means that C2 _j is located on the left side of the first acupoint corresponding to C1 _je , h2=4 means that C2 _j is located The right side of the first acupoint corresponding to C1 _je , PX _je is the number of horizontal inches between the h1 side of the first acupoint corresponding to C2 _j and C1 _je , PY _je is the h1th side of the first acupoint corresponding to C2 _j and C1 _je The number of inches between sides h2, e ranges from 1 to n1, and j ranges from 1 to n2. In this embodiment of the present invention, the ID of an acupoint may be the name of the acupoint or any other identifier capable of knowing the name of the acupoint.

In the embodiment of the present invention, a horizontal inch refers to a measurement unit between acupuncture points located in the transverse direction of the human body. For example, acupoint A is located X horizontal inches to the left or right of acupuncture point B. A straight cun refers to a unit of measurement between acupuncture points located in the longitudinal direction of the human body. For example, acupoint A is located X straight cun above or below acupoint B.

In the embodiment of the present invention, the first acupoint associated with C2 _j may be the first acupoint with the shortest distance from C2 _j . For example, the first acupoint associated with the second acupoint in the chest area is located in the chest section and the abdominal area. The first of the second associated acupoints is located in the abdominal segment and can be determined based on the existing human acupoint map.

Those skilled in the art know that n set acupoints can be obtained in the target infrared image through the first acupoint position relationship table and the second acupoint position relationship table. In practical applications, when the target infrared image is a frontal image of the human body, all the acupoints located on the Ren meridian can be obtained through the first acupoint position relationship table based on Tiantu or Shenque, and then based on the positions of the Ren meridian and the second acupoint. Through the relationship table, other acupoints, namely abdominal meridian acupoints, are obtained. When the target infrared image is an image of the back of the human body, all the acupoints located on the Du meridian can be obtained through the first acupoint position relationship table based on Dazhui or Changqiang, and then other acupoints can be obtained based on the Du meridian and the second acupoint position relationship table. The acupuncture points are the acupoints on the meridians on the back.

In some special circumstances, such as when the waistline of the human body is too thin, acupuncture points close to the human body contour may fall outside the human body contour during calculation. Therefore, the abscissa calculated in this case needs to be adjusted so that the calculated acupoints The location is within the contours of the human body.

Specifically, in S408, when acquiring any second acupuncture point C2 _j , if h2=3, if (x _je -P _je *△x)<x _c1 , then the coordinates (x _c1 +a, y _je ) As the coordinates of the acupuncture point corresponding to C2 _j ; if h2=4, if (x _je +P _je *△x)>x _c2 , then use the coordinates (x _c2 -a, y _je ) as the coordinates of the acupoint corresponding to C2 _j ; Among them, x _je and y _je are the abscissa and ordinate of the first acupoint corresponding to C1 _je respectively, x _c1 is the abscissa of the pixel point in PL whose ordinate is y _je and is located on the left side of the reference acupoint, x _c2 is PL The middle ordinate is y _je and the abscissa of the pixel point located on the right side of the reference acupuncture point. a is the preset number of pixel points. In one example, a=1.

S500, based on the obtained n set acupuncture points, divide the trunk area surrounded by the contour line of the trunk into F regions of interest, wherein the area of interest located at the outermost side of the trunk area includes the corresponding trunk contour line, that is, R1 ∩R2∩…∩Rp∩…∩RF=Null, R1∪R2∪…∪Rp∪…∪RF=R, Rp is the p-th region of interest among the F regions of interest, and R is the torso in the target infrared image corresponding area.

In the embodiment of the present invention, the F regions of interest can be set based on actual needs. Each region may include one acupuncture point or multiple acupoints. The present invention is not particularly limited as long as the F regions of interest can cover the entire trunk area, that is, the boundaries of the regions of interest located on both sides of the trunk are Human body contour lines.

The method of obtaining the region of interest provided by the embodiment of the present invention can extend at least the region of interest of the trunk part to the human body contour line, so that as many regions of interest as possible can be obtained, thereby making the analysis more accurate.

Further, in this embodiment of the present invention, the following steps are included after S200:

S210, correct m regions of interest.

In the embodiment of the present invention, each of the m regions of interest is formed by two straight lines and two limb contour lines located between the two straight lines. That is, during the annotation process of the sample image, the limbs can be Mark two straight lines up and down at the position corresponding to the area of interest. The two straight lines can be parallel lines or not. They can be set based on actual needs.

Further, in this embodiment of the present invention, S210 specifically includes:

S211, for the k-th region of interest, obtain the center O _k = (x _k , y _k ) of the k-th region of interest and obtain the relationship between the two parallel lines corresponding to the k-th region of interest and the corresponding limb outline. The coordinates of the four intersection points P _Lk1 , P _Lk2 , P _Lk3 , P _Lk4 , P _Lk1 , P _Lk2 , P _Lk3 , and P _Lk4 are respectively (x ^c _Lk1 , y ^c _Lk1 ), (x ^c _Lk2 , y ^c _Lk2 ), (x ^c _Lk3 , y ^c _Lk3 ), (x ^c _Lk4 , y ^c _Lk4 ), among them, if x _k > x _A1 , that is, the left limb, left hand or left leg, then x ^c _Lk1 ≤ ^{x c} _Lk2 < x _k , y ^c _Lk1 > y ^c _Lk2 > y _k , x ^c _Lk4 ≤ ^{x c} _Lk3 < x _k , y ^c _Lk3 > y ^c _Lk4 > y _k ; if x _k < x _A1 , it is the right limb, the right hand or Right leg, then x ^c _Lk2 ≤ ^{x c} _Lk1 < x _k , y ^c _Lk1 > y ^c _Lk2 > y _k , x ^c _Lk3 ≤ ^{x c} _Lk4 < x _k , y ^c _Lk3 > y ^c _Lk4 > y _k .

In the embodiment of the present invention, _Ok can be obtained based on existing methods. For example, x _k =Avg (x _k1 , x _k2 , ..., x _kg , ..., x _{kf (k)} ), y _k =Avg (y _k1 , y _k2 , ..., y _kg , ..., y _{kf (k)} ), (x _kg , y _kg ) are the coordinates of PROI _kg .

In the embodiment of the present invention, P _Lk1 , P _Lk2 , P _Lk3 , and P _Lk4 can be recognized through a trained image recognition model.

S212, if the straight line L1 passing through P _Lk1 and P _Lk2 and the straight line L2 passing through P _Lk3 and P _Lk4 are both located outside the k-th area of interest, obtain x _k1 =max (x ¹ _Lk , x ² _Lk ,..., x ^u1 _Lk ,..., x ^W1 _Lk ,) and x _k2 =min (x ¹ _Lk , x ² _Lk ,..., x ^u2 _Lk ,..., x ^W2 _Lk ), execute S213; if L1 and L2 are both located in the kth sense _Inside the area _{of interest} , _execute ^S214 _{; The number} of _abscissas in _{the coordinates} ^; The number of abscissas in the corresponding coordinates.

In the embodiment of the present invention, the coordinates corresponding to the limb contour line between P _Lk1 and P _Lk2 and the coordinates corresponding to the limb contour line between P _Lk3 and P _Lk4 can be obtained by PROI∩PL.

In the embodiment of the present invention, since the coordinates of P _Lk1 and P _Lk2 are known, the straight line equation of L1 can be obtained, and then all ordinates in the coordinates corresponding to the limb contour line between P _Lk1 and P _Lk2 are By substituting into the straight line equation of L1, all abscissas on L1 can be obtained, so that the coordinates corresponding to all pixels on L1 can be obtained. In the same way, the coordinates of all pixels on L2 can be obtained.

For any pixel point on L1, if the abscissa coordinate of the pixel point is smaller than the corresponding abscissa coordinate on the corresponding limb contour line, it means that L1 is located outside the corresponding limb contour line; otherwise, it is located on the corresponding limb contour line. internal. In the same way, for any pixel on L2, if the abscissa of the pixel is greater than the corresponding abscissa on the corresponding limb contour, it means that L2 is located outside the corresponding limb contour, otherwise, it is located on the corresponding limb. Inside the outline.

S213, obtain the pixel points P _k1 and P _k2 corresponding to x _k1 and x _k2 , and obtain the first rectangular area with the straight line L0 passing through P _k1 and P _k2 as the axis of symmetry. The height of the first rectangular area is H0, H0 is the set height, the unit is pixels; execute S215. H0 can be an experience value.

Those skilled in the art know that any method of obtaining the first rectangular area with the straight line L0 passing through P _k1 and P _k2 as the axis of symmetry falls within the protection scope of the present invention.

S214, obtain L = min (L1, L2, L3, L4), where L3 is a straight line passing through P _Lk2 and P _Lk3 , L4 is a straight line passing through P _Lk1 and P _Lk4 , and obtain the second line with L as the base. A rectangular area, the height of the second rectangular area is H0; execute S215.

Those skilled in the art know that any method of obtaining the second rectangular area with L as the base falls within the protection scope of the present invention. For example, based on the coordinates of the two endpoints corresponding to L and H0, the four vertex coordinates of the second rectangular area can be obtained. Specifically, for example, if L is L1, the coordinates of the two vertices of the second rectangular area can be respectively (x ^c _Lk1 , y ^c _Lk1 ), (x ^c _Lk2 , y ^c _Lk2 ), the coordinates of the other two vertices are (x ^c _Lk1 +H0, y ^c _Lk1 ), (x ^c _Lk2 +H0, y ^c _Lk2 ), if L is L2, L3, or L4, the four vertex coordinates of the second rectangular area can be obtained in a similar way to L=L1.

S215: Use the rectangular area as the modified k-th area of interest.

Those skilled in the art know that all pixels in a rectangular area can be obtained based on existing technologies. For example, the width and height of the rectangle are obtained through the four vertex coordinates of the rectangular area, and then each pixel coordinate is obtained through two nested loops. Point acquisition covers the entire rectangular area, thereby obtaining all pixels in the rectangular area.

The technical effect of S210 is that it can make the area of interest in the limbs more accurate, and in practical applications, when analyzing through temperature, the temperature of the area of interest is more accurate.

Further, in another embodiment of the present invention, S402 is replaced with:

S403, obtain the two symmetrical pixel point coordinates (x ₀₁ , y _A1 ) and (x ₀₂ , y _A1 ) with the ordinate y _A1 from PL, that is, the two intersection points of the first intersection line and the human body contour line; where, x ₀₁ <x _A1 <x ₀₂ , x ₀₃ <x _A2 <x ₀₄ .

S404 is replaced with:

S405, use (x ₀₁ , y _A1 ) or (x ₀₂ , y _A1 ) as the coordinates (x _A3 , y _A3 ) of the reference feature point CA3, that is, the two intersection points of the first intersection line and the human body contour line are used as the coordinates of CA3 .

The technical effect of S403 and S405 is to prevent the waistline in the target infrared image from being too thin, resulting in inaccurate position of the shoulder point.

Further, in this embodiment of the present invention, after S200, the following steps are also included:

S220, if x _A1 ≠ x _A2 , set x _A1 = x _A2 = x ₀ , that is, if CA1 and CA2 are not on the central axis due to errors and other reasons, they need to be adjusted to the central axis to improve the accuracy of acupoint identification. sex.

Furthermore, in the embodiment of the present invention, the area of interest of the composite area can be divided based on actual needs, as long as the divided area of interest includes corresponding contour lines. In an illustrative embodiment, when the target infrared image is a frontal image of a human body, S300 may specifically include:

S301, obtain the first fixed point b1=(x _b1 , y _b1 ) and the second fixed point b2=(x _b2 , y _b2 ) in the target infrared image, as shown in Figure 3, x _b1 =x _A1 , y _b1 =y _A1 + △V, x _b2 = x _A1 , y _b2 = y _A1 + △V/2, △V is the number of preset straight inches, preferably, △V is 3 preset straight inches, that is, the Tiantu point 3 inches above. In the embodiment of the present invention, a preset vertical inch is one-seventeenth of the height between the Tiantu point and the Shenque point, which is (y _A1 - y _A2 )/17.

S302: Obtain the first horizontal line h1 and the second horizontal line h2 respectively. As shown in Figure 3, the first horizontal line is a straight line that passes through the first fixed point and has both ends intersecting the contour lines on both sides. The second horizontal line is a straight line that passes through the second fixed point and has both ends intersecting the contour lines on both sides.

S303, obtain the first vertical line v1 and the fourth horizontal line h4. As shown in Figure 3, the first vertical line v1 is a straight line that passes through the reference point b and intersects the third horizontal line h3 and the top of the head contour line at both ends respectively. The reference point b is the end point of the first horizontal line, and the third horizontal line passes through (x _A1 , y _A1 ) and intersects the contour lines on both sides; the fourth horizontal line is the straight line passing through the intersection of the first vertical line and the contour line of the top of the head.

S304, respectively obtain the fifth to seventh horizontal lines that intersect with the contour line of the composite area, and the distance between the fifth horizontal line h5 and the fourth horizontal line h4, the distance between the fifth horizontal line h5 and the sixth horizontal line h6, the distance between the sixth horizontal line h5 and the sixth horizontal line h6. The distance between the horizontal line h6 and the seventh horizontal line h7 and the distance between the seventh horizontal line h7 and the first horizontal line h1 are equal, that is, the first vertical line is divided into four equal parts, and horizontal lines are drawn respectively at the equal dividing points to obtain the fifth to fifth Seven horizontal lines.

S305, obtain the second vertical line and the third vertical line respectively, the two ends of the second vertical line and the third vertical line are connected to the fourth horizontal line and the first horizontal line respectively, and between the second vertical line and the first vertical line The distance is equal to the distance between the end points of the second vertical line and the corresponding sixth horizontal line, and the distance between the third vertical line and the first vertical line is equal to the distance between the end points of the third vertical line and the corresponding sixth horizontal line. The distance is equal, that is, the sixth horizontal line is divided into four equal parts, and vertical lines are drawn at equally divided points on both sides to obtain the second vertical line and the third vertical line.

S306: Obtain the H regions of interest based on the first to seventh horizontal lines and the first to third vertical lines, such as the grid regions shown in Figure 3.

Further, in another embodiment of the present invention, when the target infrared image is the back image of the human body,

S300 can be replaced with:

S310, identify the Taodao point and Zhiyang point in the target infrared image, and divide the composite area surrounded by the contour lines of the head and neck into H regions of interest, wherein the region of interest located at the outermost side of the composite region includes the contour line of the corresponding composite region.

In this embodiment, the interest area division of the composite area is basically the same as S301 to S306. The difference is that in this embodiment, the acupoint for obtaining the fixed point is Taodao point, and the preset dimensions are Taodao point and Zhiyang point. One-tenth of the height between holes.

Further, the following steps are included:

S600: Visually display the acquired target infrared images of (F+m) regions of interest.

Those skilled in the art know that any method of visually displaying the acquired target infrared images of (F+m) regions of interest falls within the protection scope of the present invention. Embodiments of the present invention also provide a non-transitory computer-readable storage medium, which can be disposed in an electronic device to store at least one instruction or at least a program related to implementing a method in the method embodiments. , the at least one instruction or the at least one program section is loaded and executed by the processor to implement the method provided by the above embodiment.

An embodiment of the present invention also provides an electronic device, including a processor and the aforementioned non-transitory computer-readable storage medium.

Embodiments of the present invention also provide a computer program product, which includes program code. When the program product is run on an electronic device, the program code is used to cause the electronic device to execute the steps described above in this specification according to the present invention. steps in the method of an exemplary embodiment.

Although some specific embodiments of the invention have been described in detail by way of examples, those skilled in the art will understand that the above examples are for illustration only and are not intended to limit the scope of the invention. It will also be understood by those skilled in the art that various modifications may be made to the embodiments without departing from the scope and spirit of the invention. The scope of the present disclosure is defined by the appended claims.

Claims (10)

1. A method for obtaining a region of interest, characterized in that the method includes the following steps: S100, identify the human body contour line in the target infrared image, and obtain the corresponding human body contour line pixel point set PL = (PL ₁ , PL ₂ , ..., PL _r , ..., PL _M ), PL _r is the PL th r pixels, r ranges from 1 to M, M is the number of pixels in PL, and the coordinates of PL _r are (x _Lr , y _Lr ); S200, identify m regions of interest in the limbs in the target infrared image and the first characteristic acupuncture point CA1 and the second characteristic acupoint CA2 in the trunk, respectively, and obtain m regions of interest pixel point sets and first characteristic pixels respectively. Point PA1 and the second feature pixel point PA2; among them, the k-th region of interest pixel point set is PROI _k = (PROI _k1 , PROI _k2 ,..., PROI _kg ,..., PROI _kf(k) ), PROI _kg is PROI The g-th pixel in _k , the value of g ranges from 1 to f(k), f(k) PROI The number of pixels in _k , the value of k ranges from 1 to m; CA1 and CA2 are respectively the Among the n set acupuncture points, CA1 and CA2 are located on the central axis of the human body, CA2 is located below CA1, and the pixel coordinates of PA1 and PA2 are (x _A1 , y _A1 ) and (x _A2 , y _A2) respectively. ); S300, based on CA1 and CA2 and the set division rules, divide the composite area surrounded by the contour lines of the head and neck into H areas of interest, where the area of interest located at the outermost side of the composite area includes the corresponding composite area. contour line; S400, based on the identified CA1 and CA2, obtain the remaining acupoints among the n set acupoints in the target infrared image; S500: Divide the trunk area surrounded by the contour line of the trunk into F regions of interest based on the acquired n set acupuncture points, wherein the area of interest located at the outermost side of the trunk area includes the corresponding trunk contour line. 2. The method according to claim 1, characterized in that, after S200, it further includes the following steps: S210, correct m regions of interest; S210 specifically includes: S211, for the k-th region of interest, obtain the center of the k-th region of interest O _k = (x _k , y _k ) and obtain the relationship between the two straight lines corresponding to the k-th region of interest and the corresponding limb outline. The coordinates of the four intersection points P _Lk1 , P _Lk2 , P _Lk3 , P _Lk4 , P _Lk1 , P _Lk2 , P _Lk3 , and P _Lk4 are respectively (x ^c _Lk1 , y ^c _Lk1 ), (x ^c _Lk2 , y ^c _Lk2 ), (x ^c _Lk3 , y ^c _Lk3 ), (x ^c _Lk4 , y ^c _Lk4 ), among which, if x _k >x _A1 , then x ^c _Lk1 ≤ x ^c _Lk2 <x _k , y ^c _Lk1 > y ^c _Lk2 ＞y _k , x ^c _Lk4 ≤x ^c _Lk3 ＜x _k , y ^c _Lk3 ＞y ^c _Lk4 ＞y _k ; if x _k ＜x _A1 , then x ^c _Lk2 ≤x ^c _Lk1 ＜x _k , y ^c _Lk1 ＞y ^c _Lk2 ＞y _k , x ^c _Lk3 ≤x ^c _Lk4 ＜x _k , y ^c _Lk3 ＞y ^c _Lk4 ＞y _k ; S212, if the straight line L1 passing through P _Lk1 and P _Lk2 and the straight line L2 passing through P _Lk3 and P _Lk4 are both located outside the k-th area of interest, obtain x _k1 =max (x ¹ _Lk , x ² _Lk ,..., x ^u1 _Lk ,..., x ^W1 _Lk ,) and x _k2 =min (x ¹ _Lk , x ² _Lk ,..., x ^u2 _Lk ,..., x ^W2 _Lk ), execute S213; if L1 and L2 are both located in the kth sense _Inside the area _{of interest} , _execute ^S214 _{; The number} of _{abscissas in the} coordinates ^; The number of abscissas in the corresponding coordinates; S213, obtain the pixel points P _k1 and P _k2 corresponding to x _k1 and x _k2 , and obtain the first rectangular area with the straight line L0 passing through P _k1 and P _k2 as the axis of symmetry. The height of the first rectangular area is H0, H0 is the set height, unit is pixel; execute S215; S214, obtain L = min (L1, L2, L3, L4), where L3 is a straight line passing through P _Lk2 and P _Lk3 , L4 is a straight line passing through P _Lk1 and P _Lk4 , and obtain the second line with L as the base. A rectangular area, the height of the second rectangular area is H0; execute S215; S215: Use the rectangular area as the modified k-th area of interest. 3. The method according to claim 1, characterized in that S500 specifically includes: S402, obtain the coordinates (x ₀₁ , y _{A1 )} and (x ₀₂ , y _A1 ) of two symmetrical pixels with the vertical coordinate y A1 from PL, and obtain the two symmetrical pixels with the vertical coordinate y _A2 from PL. Coordinates (x ₀₃ , y _A2 ) and (x ₀₄ , y _A2 ); among them, x ₀₁ <x _A1 <x ₀₂ , x ₀₃ <x _A2 <x ₀₄ ; S404, obtain the pixel coordinates corresponding to (min(x ₀₁ , x ₀₃ ), y _A1 ) or (min(x ₀₂ , x ₀₄ ), y _A1 ) as the coordinates (x _A3 , y _A3 ) of the reference feature point CA3; S406, obtain the horizontal length △x＝∣x _A3 -x _A1 ∣/W ₁₃ , and obtain the vertical length △y＝∣y _A2 -y _A1 ∣/V ₁₂ , W ₁₃ is the horizontal length between CA1 and CA3 Quantity, V ₁₂ is the number of inches between CA1 and CA2; S408: Using CA1 or CA2 as the reference acupoint, based on Δx and Δy and the position relationship table between the n set acupuncture points, obtain the remaining acupoints among the n set acupuncture points in the target infrared image. 4. The method according to claim 3, characterized in that S402 is replaced by: S403, obtain the coordinates (x ₀₁ , y _A1 ) and (x ₀₂ , y _A1 ) of two symmetrical pixels whose ordinates are y _A1 from PL; among them, x ₀₁ ＜x _A1 ＜x ₀₂ , x ₀₃ ＜x _A2 <x ₀₄ ; S404 is replaced with: S405, use (x ₀₁ , y _A1 ) or (x ₀₂ , y _A1 ) as the coordinates (x _A3 , y _A3 ) of the reference feature point CA3; CA3 is an acupoint among the n set acupoints. 5. The method according to claim 3, wherein the n set acupuncture points include n1 first acupoints located on the central axis of the human body and n2 second acupoints located on both sides of the central axis of the human body, and the The position relationship table between the n set acupuncture points includes a first acupoint position relationship table and a second acupoint position relationship table, where the i-th row of the first acupoint position relationship table includes (C1 _i , h1, P _ci ), C1 _i is the ID of the i-th first acupoint among the n1 first acupoints, i ranges from 1 to n1, h1 takes a value of 1 or 2, where h1=1 means that C1 _i is located on the upper side of the reference acupoint , h1=2 means that C1 _i is located on the lower side of the reference acupoint, P _ci means the number of inches between the i-th first acupoint and the reference acupoint; the j-th row of the second acupoint position relationship table includes (C2 _j , C1 _je , (h1, PX _je ), (h2, PY _je )), C2 _j is the ID of the j-th second acupoint among the n2 second acupoints, C1 _je is the ID of the first acupoint associated with C2 _j , h2 The value of is 3 or 4, h2=3 means that C2 _j is located on the left side of the first acupoint corresponding to C1 _je , h2=4 means that C2 _j is located on the right side of the first acupoint corresponding to C1 _je , PX _je is C2 _j and C1 _je is the number of horizontal inches between the h1th side of the first acupuncture point corresponding to C1 je, PY _je is the number of vertical inches between the h2th side of the first acupuncture point corresponding to C2 _j and C1 _je , and the value of e is 1 to n1 , the value of j is from 1 to n2. 6. The method according to claim 5, characterized in that, in S408, when acquiring any second acupoint _C2j , if h2=3, if (xje _{- Pje} *Δx)< _xc1 , Then use the coordinates (x _c1 +a, y _je ) as the coordinates of the acupuncture point corresponding to C2 _j ; if h2=4, if (x _je +P _je *△x)>x _c2 , then use the coordinates (x _c2 -a, y _je ) as the coordinates of the acupoint corresponding to C2 _j ; among them, x _je and y _je are respectively the abscissa and ordinate coordinates of the first acupoint corresponding to C1 _je , x _c1 is the ordinate in PL with y _je and is located to the left of the reference acupoint The abscissa coordinate of the pixel point on the side, x _c2 is the abscissa coordinate of the pixel point in PL with the ordinate coordinate y _je and located on the right side of the reference acupuncture point, a is the preset number of pixel points. 7. The method according to claim 1, characterized in that, PL is obtained based on the following steps: S101, identify the human body contour in the target infrared image, and obtain the corresponding human body contour initial pixel point set PLI = (PLI ₁ , PLI ₂ , ..., PLI _r , ..., PLI _M ); S102, obtain the temperature set corresponding to PLI TCI = (TLI ₁ , TLI ₂ , ..., TLI _r , ..., TLI _M ), where TLI _r is the temperature corresponding to PLI _r ; S103, if TLI _r <T0, execute S104; T0 is the preset temperature value; otherwise, execute S106; S104, if x _LIr <x ₀ , set x _LIr = x _LIr + b, if x _LIr > x ₀ , set x _LIr = x _LIr -b; x ₀ is the abscissa of the central axis, x _LIr is TLI _r The abscissa of the corresponding pixel, b is the preset number of pixels; execute S105; S105, if the temperature corresponding to the pixel point (x _LIr , y _LIr ) is less than T0, execute S104; otherwise, execute S106; S106, set r=r+1, if r≤M, execute S103; otherwise, exit the control program. 8. The method according to claim 1, characterized in that the target infrared image is a frontal image of the human body; S300 specifically includes: S301, obtain the first fixed point b1=(x _b1 , y _b1 ) and the second fixed point b2=(x _b2 , y _b2 ), x _b1 =x _A1 , y _b1 =y _A1 + △V in the target infrared image , x _b2 = x _A1 , y _b2 = y _A1 + △V/2, △V is the number of preset straight inches; among them, 1 preset straight size = (y _A1 - y _A2 )/17; S302, obtain the first horizontal line and the second horizontal line respectively. The first horizontal line is a straight line that passes through the first fixed point and both ends intersect with the contour lines on both sides. The second horizontal line is a straight line that passes through the second fixed point and both ends intersect with the contour lines on both sides. intersecting straight lines; S303. Obtain the first vertical line and the fourth horizontal line. The first vertical line is a straight line that passes through the reference point and intersects the third horizontal line and the top of the head contour line at both ends respectively. The reference point is the end point of the first horizontal line. The third horizontal line passes through (x _A1 , y _A1 ) and intersects the contour lines on both sides; the fourth horizontal line is the straight line passing through the intersection of the first vertical line and the top contour line; S304, respectively obtain the fifth to seventh horizontal lines that intersect with the contour line of the composite area, and the distance between the fifth horizontal line and the fourth horizontal line, the distance between the fifth horizontal line and the sixth horizontal line, the sixth horizontal line and the seventh horizontal line. The distance between the horizontal lines and the distance between the seventh and first horizontal lines are equal; S305, obtain the second vertical line and the third vertical line respectively, the two ends of the second vertical line and the third vertical line are connected to the fourth horizontal line and the first horizontal line respectively, and between the second vertical line and the first vertical line The distance is equal to the distance between the end points of the second vertical line and the corresponding sixth horizontal line, and the distance between the third vertical line and the first vertical line is equal to the distance between the end points of the third vertical line and the corresponding sixth horizontal line. equal distance; S306: Obtain the H regions of interest based on the first to seventh horizontal lines and the first to third vertical lines. 9. A non-transitory computer-readable storage medium, in which at least one instruction or at least one program is stored, characterized in that the at least one instruction or the at least one program is loaded and executed by a processor. Implement the method according to any one of claims 1-8. 10. An electronic device, characterized by comprising a processor and the non-transitory computer-readable storage medium described in claim 9.