CN109241878B - Lip positioning-based facial feature positioning method and system - Google Patents

Abstract

The invention provides a facial feature positioning method and system based on lip positioning. Firstly, finding a lip area of a current face, and determining face inclination and face side degree; then, determining a search area of eyes and noses according to the inclination and the side degree of the face, and then accurately judging according to the characteristics of the eyes and the noses to complete the positioning of the five sense organs of the current face. Reasonable inclination angle analysis can improve the accuracy of face recognition on one hand, and on the other hand, an accurate positioning search direction can be provided for an incomplete eye and nose positioning system.

Description

Lip positioning-based facial feature positioning method and system

Technical Field

The invention relates to the technical field of image processing, in particular to a lip positioning-based facial feature positioning method and system.

Background

Face recognition and online video beauty are two emerging video applications. In practical application, the human face is not always in the front and does not skew. Due to personal habits and needs for beautification, there are often situations such as sideways, leaning, etc. If the face is not distinguished according to the normal mode, the accuracy of face detection and recognition can be affected. The human face detection is inaccurate, and the good portrait beautifying effect cannot be achieved naturally.

Disclosure of Invention

The embodiment of the invention aims to provide a facial feature positioning method based on lip positioning, and aims to solve the problem of low face detection and identification accuracy in the prior art.

The embodiment of the invention is realized by a lip positioning-based facial feature positioning method, which comprises the following steps:

step 1: performing lip positioning on the current face;

step 2: finding a block column with the minimum column number in the face lip block set judged to be the current face lip block set, and positioning a middle block of the block column and marking as gmb (i1, j 1); finding the block column with the largest column number, locating the middle block of the block column, denoted gmb (i2, j 2);

step 3: finding a block row with the minimum row number in the face lip block set judged to be current, and positioning a middle block of the block row, and marking as gmb (i3, j 3); finding the block row with the largest row number, locating the middle block of the block row, and recording as gmb (i4, j 4);

step 4: calculating the current face inclination angle theta;

step 5: calculating the side proportion gamma of the current face;

step 6: determining an eye-nose to-be-detected area of the current face according to the face inclination angle and the face side ratio;

step 7: and carrying out accurate eye-nose judgment on the blocks in the eye-nose to-be-detected region of the current face to complete the positioning of the five sense organs of the current face.

It is another object of an embodiment of the present invention to provide a facial feature positioning system based on lip positioning, the system including:

the lip positioning device is used for carrying out lip positioning on the current face;

a first middle block searching and positioning module, which finds a block column with the minimum column number in the set of the lip blocks of the face judged to be current, positions the middle block of the block column and records as gmb (i1, j 1); finding the block column with the largest column number, locating the middle block of the block column, denoted gmb (i2, j 2);

a second intermediate block searching and positioning module, which finds the block row with the minimum row number in the set of the face lip blocks judged to be current, positions the intermediate block of the block row and records as gmb (i3, j 3); finding the block row with the largest row number, locating the middle block of the block row, and recording as gmb (i4, j 4);

the face inclination angle calculation module is used for calculating the current face inclination angle theta;

the face side proportion calculation module is used for calculating the current face side proportion gamma;

the eye-nose to-be-detected region determining module is used for determining the eye-nose to-be-detected region of the current face according to the face inclination angle and the face side proportion;

and the eye and nose accurate judgment module is used for accurately judging the eye and nose of the current face in the eye and nose to-be-detected region, so as to complete the positioning of the five sense organs of the current face.

The invention has the advantages of

The invention provides a facial feature positioning method and system based on lip positioning. Firstly, finding a lip area of a current face, and determining face inclination and face side degree; then, determining a search area of eyes and noses according to the inclination and the side degree of the face, and then accurately judging according to the characteristics of the eyes and the noses to complete the positioning of the five sense organs of the current face. Reasonable inclination angle analysis can improve the accuracy of face recognition on one hand, and on the other hand, an accurate positioning search direction can be provided for an incomplete eye and nose positioning system.

Drawings

FIG. 1 is a flow chart of a method for positioning facial features based on lip positioning according to a preferred embodiment of the present invention;

fig. 2 is a diagram of a facial feature positioning system based on lip positioning according to a preferred embodiment of the present invention.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail below with reference to the accompanying drawings and examples, and for convenience of description, only parts related to the examples of the present invention are shown. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

The invention provides a facial feature positioning method and system based on lip positioning. Firstly, finding a lip area of a current face, and determining face inclination and face side degree; then, determining a search area of eyes and noses according to the inclination and the side degree of the face, and then accurately judging according to the characteristics of the eyes and the noses to complete the positioning of the five sense organs of the current face. Reasonable inclination angle analysis can improve the accuracy of face recognition on one hand, and on the other hand, an accurate positioning search direction can be provided for an incomplete eye and nose positioning system.

Example one

FIG. 1 is a flow chart of a method for positioning facial features based on lip positioning according to a preferred embodiment of the present invention; the method comprises the following steps:

step 1: and carrying out lip positioning on the current face.

The lip positioning method employs methods disclosed in the art.

Step 2: finding a block column with the minimum column number in the face lip block set judged to be the current face lip block set, and positioning a middle block of the block column and marking as gmb (i1, j 1); the block column with the largest column number is found, and the middle block of the block column is located, denoted gmb (i2, j 2).

Step 3: finding a block row with the minimum row number in the face lip block set judged to be current, and positioning a middle block of the block row, and marking as gmb (i3, j 3); the block row with the largest row number is found, and the middle block of this block row is located, denoted gmb (i4, j 4).

Step 4: calculating the inclination angle theta of the current face,

step 5: calculating the side proportion gamma of the current face,

step 6: and determining the eye-nose to-be-detected region of the current face according to the face inclination angle and the face side ratio.

Case1：θ＝0

If gamma is a sufficient right-side face, only a single-side human eye judgment region exists, and the single-side human eye judgment region is defined as { gmb (i, j) |2 × i3-i4-Thres_ie≤i≤i3-Thres_ieAnd j1-Thres_jeJ is not less than (j3+ j4)/2}, and the nose judgment area is defined as { gmb (i, j) |2 × i3-i4 is not less than i3 and j1 is not less than j is not less than (j3+ j4)/2 };

otherwise, if gamma is a sufficient left-side face, only a single-side human eye judgment region exists, and the single-side human eye judgment region is defined as { gmb (i, j) |2 × i3-i4-Thres_ie≤i≤i3-Thres_ieAnd (j3+ j4)/2 is not less than j2+ Thres_jeDefining a nose judgment area as { gmb (i, j) |2 × i3-i4 ≤ i ≤ i3 and (j3+ j4)/2 ≤ j ≤ j2 };

otherwise, demarcating the human eye judgment area as

{gmb(i,j)|2*i3-i4-Thres_ie≤i≤i3-Thres_ieAnd j1-Thres_jL≤j≤j2+Thres_jRAnd defining a nose judgment area as { gmb (i, j) |2 i3-i4 ≦ i3 and j1 ≦ j2 }.

Wherein Thres < Thres_ie＜3*Thres，Thres＝max{i4-i3,j2-j1}，

0.5*Thres＜Thres_je＜2*Thres、

Thres、Thres_ie、Thres_je、Thres_jL、Thres_jRRespectively called threshold, first threshold, second threshold, left column threshold, right column threshold; solving the maximum value by a max table; gmb (i, j) denotes a block located at ith row and jth column of the current picture. .

Case2:θ≠0

Firstly, defining a rectangular area of an eye nose according to corresponding gamma in Case 1; then finding respective central points of the eye rectangular region and the nose rectangular region, and respectively rotating the eye-nose rectangular region by theta by taking the respective central points as rotating fixed points; and finally, selecting the blocks falling into the rotating rectangular area as the eye-nose rectangular area. Step 7: and carrying out accurate eye-nose judgment on the blocks in the eye-nose to-be-detected region of the current face to complete the positioning of the five sense organs of the current face.

The method disclosed in the industry is adopted for accurately judging the eye nose of the current human face in the eye nose to-be-detected area.

Example two

Fig. 2 is a diagram of a facial feature positioning system based on lip positioning according to a preferred embodiment of the present invention. The system comprises:

and the lip positioning device is used for carrying out lip positioning on the current face.

The lip positioning of the current face adopts an industry-disclosed method;

a first middle block searching and positioning module, which finds a block column with the minimum column number in the set of the lip blocks of the face judged to be current, positions the middle block of the block column and records as gmb (i1, j 1); the block column with the largest column number is found, and the middle block of the block column is located, denoted gmb (i2, j 2).

A second intermediate block searching and positioning module, which finds the block row with the minimum row number in the set of the face lip blocks judged to be current, positions the intermediate block of the block row and records as gmb (i3, j 3); the block row with the largest row number is found, and the middle block of this block row is located, denoted gmb (i4, j 4).

A face inclination angle calculation module for calculating the current face inclination angle theta,

a face side proportion calculating module for calculating the current face side proportion gamma,

and the eye-nose to-be-detected region determining module is used for determining the eye-nose to-be-detected region of the current face according to the face inclination angle and the face side proportion.

Case1：θ＝0

If gamma is a sufficient right-side face, only a single-side human eye judgment region exists, and the single-side human eye judgment region is defined as { gmb (i, j) |2 × i3-i4-Thres_ie≤i≤i3-Thres_ieAnd j1-Thres_jeJ is not less than (j3+ j4)/2}, and the nose judgment area is defined as { gmb (i, j) |2 × i3-i4 is not less than i3 and j1 is not less than j is not less than (j3+ j4)/2 };

otherwise, if gamma is a sufficient left-side face, only a single-side human eye judgment region exists, and the single-side human eye judgment region is defined as { gmb (i, j) |2 × i3-i4-Thres_ie≤i≤i3-Thres_ieAnd (j3+ j4)/2 is not less than j2+ Thres_jeDefining a nose judgment area as { gmb (i, j) |2 × i3-i4 ≤ i ≤ i3 and (j3+ j4)/2 ≤ j ≤ j2 };

otherwise, demarcating the human eye judgment area as

{gmb(i,j)|2*i3-i4-Thres_ie≤i≤i3-Thres_ieAnd j1-Thres_jL≤j≤j2+Thres_jRAnd defining a nose judgment area as { gmb (i, j) |2 i3-i4 ≦ i3 and j1 ≦ j2 }.

Wherein Thres < Thres_ie＜3*Thres，Thres＝max{i4-i3,j2-j1}，

0.5*Thres＜Thres_je＜2*Thres、

Thres、Thres_ie、Thres_je、Thres_jL、Thres_jRRespectively called threshold, first threshold, second threshold, left column threshold, right column threshold; solving the maximum value by a max table; gmb (i, j) denotes a block located at ith row and jth column of the current picture.

Case2:θ≠0

Firstly, defining a rectangular area of an eye nose according to corresponding gamma in Case 1; then finding respective central points of the eye rectangular region and the nose rectangular region, and respectively rotating the eye-nose rectangular region by theta by taking the respective central points as rotating fixed points; and finally, selecting the blocks falling into the rotating rectangular area as the eye-nose rectangular area.

And the eye and nose accurate judgment module is used for accurately judging the eye and nose of the current face in the eye and nose to-be-detected region, so as to complete the positioning of the five sense organs of the current face.

The accurate eye-nose judgment of the block in the eye-nose to-be-detected area of the current face adopts an industry-disclosed method.

It will be understood by those skilled in the art that all or part of the steps in the method according to the above embodiments may be implemented by hardware related to program instructions, and the program may be stored in a computer readable storage medium, such as ROM, RAM, magnetic disk, optical disk, etc.

The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents and improvements made within the spirit and principle of the present invention are intended to be included within the scope of the present invention.

Claims (4)

1. A method for lip location based facial features localization, the method comprising:

step 1: performing lip positioning on the current face;

step 2: finding a block column with the minimum column number in the face lip block set judged to be the current face lip block set, and positioning a middle block of the block column and marking as gmb (i1, j 1); finding the block column with the largest column number, locating the middle block of the block column, denoted gmb (i2, j 2);

step 3: finding a block row with the minimum row number in the face lip block set judged to be current, and positioning a middle block of the block row, and marking as gmb (i3, j 3); finding the block row with the largest row number, locating the middle block of the block row, and recording as gmb (i4, j 4);

step 4: calculating the current face inclination angle theta:

step 5: calculating the side proportion gamma of the current face:

step 6: according to the face inclination angle and the face side proportion, determining an eye-nose to-be-detected region of the current face, comprising:

Case1：θ＝0

if gamma is a sufficient right-side face, only a single-side human eye judgment region exists, and the single-side human eye judgment region is defined as { gmb (i, j) |2 × i3-i4-Thres_ie≤i≤i3-Thres_ieAnd j1-Thres_jeJ is not less than (j3+ j4)/2}, and the nose judgment area is defined as { gmb (i, j) |2 × i3-i4 is not less than i3 and j1 is not less than j is not less than (j3+ j4)/2 };

otherwise, if gamma is a sufficient left-side face, only a single-side human eye judgment region exists, and the single-side human eye judgment region is defined as { gmb (i, j) |2 × i3-i4-Thres_ie≤i≤i3-Thres_ieAnd (j3+ j4)/2 is not less than j2+ Thres_jeDefining a nose judgment area as { gmb (i, j) |2 × i3-i4 ≤ i ≤ i3 and (j3+ j4)/2 ≤ j ≤ j2 };

otherwise, the human eye judgment region is defined as { gmb (i, j) |2 × i3-i4-Thres_ie≤i≤i3-Thres_ieAnd j1-Thres_jL≤j≤j2+Thres_jRDefining a nose judgment area as { gmb (i, j) |2 i3-i4 ≤ i ≤ i3 and j1 ≤ j2 };

wherein Thres < Thres_ie＜3*Thres，Thres＝max{i4-i3,j2-j1}，0.5*Thres＜Thres_je＜2*Thres、

Thres、Thres_ie、Thres_je、Thres_jL、Thres_jRRespectively called threshold, first threshold, second threshold, left column threshold, right column threshold; solving the maximum value by a max table; gmb (i, j) represents a block located at ith row and jth column of the current picture;

step 7: and carrying out accurate eye-nose judgment on the blocks in the eye-nose to-be-detected region of the current face to complete the positioning of the five sense organs of the current face.

2. The lip positioning-based facial feature positioning method according to claim 1, wherein the determining the eye-nose to-be-detected region of the current face according to the face inclination angle and the face side ratio further comprises:

Case2:θ≠0

firstly, defining a rectangular area of an eye nose according to corresponding gamma in Case 1; then finding respective central points of the eye rectangular region and the nose rectangular region, and respectively rotating the eye-nose rectangular region by theta by taking the respective central points as rotating fixed points; and finally, selecting the blocks falling into the rotating rectangular area as the eye-nose rectangular area.

3. A lip-positioning-based facial feature positioning system, comprising:

the lip positioning device is used for carrying out lip positioning on the current face;

a first middle block searching and positioning module, which finds a block column with the minimum column number in the set of the lip blocks of the face judged to be current, positions the middle block of the block column and records as gmb (i1, j 1); finding the block column with the largest column number, locating the middle block of the block column, denoted gmb (i2, j 2);

a second intermediate block searching and positioning module, which finds the block row with the minimum row number in the set of the face lip blocks judged to be current, positions the intermediate block of the block row and records as gmb (i3, j 3); finding the block row with the largest row number, locating the middle block of the block row, and recording as gmb (i4, j 4);

the face inclination angle calculation module calculates the current face inclination angle theta:

the face side proportion calculation module is used for calculating the current face side proportion gamma:

the eye-nose detection region determining module is used for determining the eye-nose detection region of the current face according to the face inclination angle and the face side proportion:

Case1：θ＝0

if gamma is a sufficient right-side face, only a single-side human eye judgment region exists, and the single-side human eye judgment region is defined as { gmb (i, j) |2 × i3-i4-Thres_ie≤i≤i3-Thres_ieAnd j1-Thres_jeJ is not less than (j3+ j4)/2}, and the nose judgment area is defined as { gmb (i, j) |2 × i3-i4 is not less than i3 and j1 is not less than j is not less than (j3+ j4)/2 };

otherwise, if gamma is a sufficient left-side face, only a single-side human eye judgment region exists, and the single-side human eye judgment region is defined as { gmb (i, j) |2 × i3-i4-Thres_ie≤i≤i3-Thres_ieAnd (j3+ j4)/2 is not less than j2+ Thres_jeDefining a nose judgment area as { gmb (i, j) |2 × i3-i4 ≤ i ≤ i3 and (j3+ j4)/2 ≤ j ≤ j2 };

otherwise, the human eye judgment region is defined as { gmb (i, j) |2 × i3-i4-Thres_ie≤i≤i3-Thres_ieAnd j1-Thres_jL≤j≤j2+Thres_jRDefining a nose judgment area as { gmb (i, j) |2 i3-i4 ≤ i ≤ i3 and j1 ≤ j2 };

wherein Thres < Thres_ie＜3*Thres，Thres＝max{i4-i3,j2-j1}，

0.5*Thres＜Thres_je＜2*Thres、

Thres、Thres_ie、Thres_je、Thres_jL、Thres_jRRespectively called threshold, first threshold, second threshold, left column threshold, right column threshold; solving the maximum value by a max table; gmb (i, j) represents a block located at ith row and jth column of the current picture;

and the eye and nose accurate judgment module is used for accurately judging the eye and nose of the current face in the eye and nose to-be-detected region, so as to complete the positioning of the five sense organs of the current face.

4. The lip-positioning-based facial feature positioning system of claim 3,

in the eye-nose detection region determining module, the eye-nose detection region determining module is further configured to determine an eye-nose detection region of the current face according to the face inclination angle and the face side ratio:

Case2:θ≠0

firstly, defining a rectangular area of an eye nose according to corresponding gamma in Case 1; then finding respective central points of the eye rectangular region and the nose rectangular region, and respectively rotating the eye-nose rectangular region by theta by taking the respective central points as rotating fixed points; and finally, selecting the blocks falling into the rotating rectangular area as the eye-nose rectangular area.

Images