CN102938139B - Automatic synthesis method for fault finding game images - Google Patents
Automatic synthesis method for fault finding game images Download PDFInfo
- Publication number
- CN102938139B CN102938139B CN201210447588.6A CN201210447588A CN102938139B CN 102938139 B CN102938139 B CN 102938139B CN 201210447588 A CN201210447588 A CN 201210447588A CN 102938139 B CN102938139 B CN 102938139B
- Authority
- CN
- China
- Prior art keywords
- region
- represent
- amendment
- modified
- difficulty
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Abstract
The invention discloses an automatic synthesis method for fault finding game images. The automatic synthesis method for the fault finding game images specifically comprises the steps of inputting images to be processed and an expected difficulty value; performing region division and object detection to input images and finding regions to be modified and parameters to be modified; utilizing the modifying difficulty to evaluate functions and performing iterative optimization to the regions to be modified and the parameters to be modified; performing image synthesis according to modified regions and modified parameters obtained through optimization; and outputting an image synthesis result. The method can automatically perform image synthesis of the fault finding game images, supports the difficulty value of expected generated images input by a user and is a first difficulty-controllable automatic synthesis method for the fault finding game images.
Description
Technical field
The present invention relates to technical field of image processing, be specifically related to the game image automatic synthesis method of finding fault that a kind of difficulty is controlled.
Background technology
Image is found fault and played is a kind of developmental game extensively liked by masses, but the image major part that game of finding fault all at present uses is all the result of artificial treatment (such as using photo handling software), and the image that minority generates automatically also has the problems such as image scene is destroyed, synthetic effect is too poor.The difficulty that image is found fault generally manually is specified based on experience.These problems game image that makes to find fault has that combined coefficient is low, man-machine interactively is many, difficulty is difficult to the defects such as appointment.In addition, game image of finding fault also can be used in that psychologic amendment is blind looks field, also has very large meaning for the efficiency and accuracy improving Experiment of Psychology.
Blindly looking in field in psychology amendment, having a lot of research work to be sum up and predict the impact of input picture on amendment difficulty by manually testing.Wherein relatively more relevant is that Rensink proposed " To see or not to see:The need for attention to perceivechanges in scenes " in 1997, this work thinks that amendment difficulty is relevant to the visual importance of modifier area, the visual importance of modifier area is stronger, and amendment difficulty is lower.Verma proposed in 2010 " A semi-automated approach to balancing of bottom-up salience forpredicting change detection performance ", this work proposes that a kind of automanual amendment is blind looks image combining method, the method by ensure modifier area before a modification after visual importance reach amendment blind object of looking, but this work still can not control revise difficulty.
Summary of the invention
(1) technical matters that will solve
The technical matters that the present invention mainly solves is: the game image of finding fault how carrying out difficulty controlled synthesizes automatically.
(2) technical scheme
For solving the problem, the invention provides the game image automatic synthesis method of finding fault that a kind of difficulty is controlled, comprising the following steps:
S1, input pending image and expect difficulty value;
S2, region segmentation and object detection are carried out to input picture, find out region to be modified and parameter to be modified;
S3, utilization amendment difficulty evaluation function carry out iteration optimization to described region to be modified and parameter to be modified;
S4, according to optimizing the modifier area that obtains and amendment parameter carries out Images uniting;
S5, output image synthesis result.
Preferably, described amendment difficulty evaluation function is:
B=exp(-max(‖I
K‖S(I
K),‖I
K′||S(I
K′))·D(I
K,I
K′)),
Wherein I
k, I
k' represent the position of the region of amendment in input picture and output image respectively;
‖ I
k‖ represents region I
kthe number of pixels comprised;
S (I
k) represent region I
kvisual importance degree;
D (I
k, I
k') represent region I
kand I
kthe index word of the retouching operation of ' correspondence.
Preferably, described visual importance degree is the product of color importance degree and background complexity, that is:
S(I
K)=S
O(I
K)C(I
K),
Wherein S
o(I
k) represent region I
kcolor importance degree;
C (I
k) represent region I
kbackground complexity.
Preferably, the index word of described retouching operation is expressed as region I
kand I
k' respectively in the weighted mean of color, texture and shape space middle distance, that is:
D(I
K,I
K′)=ω
cD
c+ω
tD
t+ω
zD
z,
Wherein D
crepresent region I
kand I
k' distance in color distribution space;
D
trepresent region I
kand I
k' distance in grain distribution space;
D
zrepresent region I
kand I
k' distance in shape space;
ω
c, ω
t, ω
zrepresent three constants.
Preferably, described iteration optimization uses if minor function is as target:
|B-B
*|
2+λmax(M
min-M,0),
Wherein B represents the amendment difficulty that current amendment is corresponding;
B
*represent the expectation difficulty value of input;
M represents the result figure that current amendment is corresponding and the difference of former figure in pixel scale;
λ and M
minrepresent two constants.
(3) beneficial effect
The inventive method can carry out the Images uniting of playing of finding fault automatically, and supports that user inputs the difficulty value expecting synthetic image.
Accompanying drawing explanation
Fig. 1 is the process flow diagram of the inventive method.
Embodiment
Below in conjunction with drawings and Examples, the specific embodiment of the present invention is described in further detail.Following examples for illustration of the present invention, but are not used for limiting the scope of the invention.
Fig. 1 is the process flow diagram of the inventive method, the invention provides one and to find fault game image automatic synthesis method, comprise the following steps:
S1: input pending image and expect difficulty value.The method can select suitable modifier area and amendment parameter in the input image according to expectation difficulty value, then is synthesized the game image pair of finding fault having and expect difficulty value by the method for Images uniting.
S2: region segmentation and object detection are carried out to the image of input, finds out region to be modified and parameter to be modified.This step is split input picture by existing image region segmentation algorithm, more therefrom finds the region to be modified of a reasonable set to carry out iteration optimization by morphology operations.
S3: utilize amendment difficulty evaluation function to carry out iteration optimization to described region to be modified and parameter to be modified.Amendment difficulty evaluation function is:
B=exp(-max(‖I
K‖S(I
K),‖I
K′‖S(I
K′))·D(I
K,I
K′))
Wherein I
k, I
k' represent the position of the region of amendment in input picture and output image respectively;
‖ I
k‖ represents region I
kthe number of pixels comprised;
S (I
k) represent region I
kvisual importance degree;
D (I
k, I
k') represent region I
kand I
kthe index word of the retouching operation of ' correspondence.
Which use the product visual importance degree as a whole of color importance degree and background complexity, that is:
S(I
K)=S
O(I
K)C(I
K)
Wherein S
o(I
k) represent region I
kcolor importance degree;
C (I
k) represent region I
kbackground complexity.
The index word of the retouching operation in amendment difficulty evaluation function is expressed as I
kand I
k' respectively in the weighted mean of color, texture and shape space middle distance, that is:
D(I
K,I
K′)=ω
cD
c+ω
tD
t+ω
zD
z
Wherein D
crepresent region I
kand I
k' distance in color distribution space;
D
trepresent region I
kand I
k' distance in grain distribution space;
D
zrepresent region I
kand I
k' distance in shape space;
ω
c, ω
t, ω
zrepresent three constants.
The iteration optimization that the method utilizes amendment difficulty evaluation function to carry out modifier area and amendment parameter, uses if minor function is as iteration optimization target:
|B-B
*|
2+λmax(M
min-M,0)
Wherein B represents the amendment difficulty that current amendment is corresponding;
B
*represent the expectation difficulty value of input;
M represents the result figure that current amendment is corresponding and the difference of former figure in pixel scale;
λ and M
minrepresent two constants.
S4: carry out Images uniting with amendment parameter according to optimizing the modifier area obtained.Iteration optimization obtains and meets modifier area and the amendment parameter that difficulty value is expected in input, and this step is revised input picture automatically by existing Images uniting algorithm.
S5: output image synthesis result.
Method of the present invention can carry out the Images uniting of playing of finding fault automatically, supports that user inputs the difficulty value expecting synthetic image.
The above is only the preferred embodiment of the present invention; it should be pointed out that for those skilled in the art, under the prerequisite not departing from the technology of the present invention principle; can also make some improvement and replacement, these improve and replace and also should be considered as protection scope of the present invention.
Claims (4)
1. to find fault a game image automatic synthesis method, it is characterized in that, comprise the following steps:
S1, input pending image and expect difficulty value;
S2, region segmentation and object detection are carried out to input picture, find out region to be modified and parameter to be modified;
S3, utilization amendment difficulty evaluation function carry out iteration optimization to described region to be modified and parameter to be modified;
S4, according to optimizing the modifier area that obtains and amendment parameter carries out Images uniting;
S5, output image synthesis result;
Described iteration optimization uses if minor function is as target:
|B-B
*|
2+λmax(M
min-M,0),
Wherein B represents the amendment difficulty that current amendment is corresponding;
B
*represent the expectation difficulty value of input;
M represents the result figure that current amendment is corresponding and the difference of former figure in pixel scale;
λ and M
minrepresent two constants.
2. the method for claim 1, is characterized in that, described amendment difficulty evaluation function is:
B=exp(-max(||I
K||S(I
K),||I
K'||S(I
K'))·D(I
K,I
K')),
Wherein I
k, I
k' be illustrated respectively in the region of the amendment in input picture, output image;
|| I
k|| represent region I
kthe number of pixels comprised;
|| I
k' || represent region I
k' the number of pixels that comprises;
S (I
k) represent region I
kvisual importance degree;
S (I
k') represent region I
k' visual importance degree;
D (I
k, I
k') represent region I
kand I
k' the index word of corresponding retouching operation.
3. method as claimed in claim 2, it is characterized in that, described visual importance degree is the product of color importance degree and background complexity, that is:
S(I
K)=S
0(I
K)C(I
K),
Wherein S
0(I
k) represent region I
kcolor importance degree;
C (I
k) represent region I
kbackground complexity.
4. method as claimed in claim 2, it is characterized in that, the index word of described retouching operation is expressed as region I
kand I
k' respectively in the weighted mean of color, texture and shape space middle distance, that is:
D(I
K,I
K')=ω
cD
c+ω
tD
t+ω
zD
z,
Wherein D
crepresent region I
kand I
k' distance in color distribution space;
D
trepresent region I
kand I
k' distance in grain distribution space;
D
zrepresent region I
kand I
k' distance in shape space;
ω
c, ω
t, ω
zrepresent three constants.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201210447588.6A CN102938139B (en) | 2012-11-09 | 2012-11-09 | Automatic synthesis method for fault finding game images |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201210447588.6A CN102938139B (en) | 2012-11-09 | 2012-11-09 | Automatic synthesis method for fault finding game images |
Publications (2)
Publication Number | Publication Date |
---|---|
CN102938139A CN102938139A (en) | 2013-02-20 |
CN102938139B true CN102938139B (en) | 2015-03-04 |
Family
ID=47697032
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201210447588.6A Active CN102938139B (en) | 2012-11-09 | 2012-11-09 | Automatic synthesis method for fault finding game images |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN102938139B (en) |
Families Citing this family (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106582025A (en) * | 2016-11-24 | 2017-04-26 | 北京暴风魔镜科技有限公司 | Game making system and method |
CN110292774B (en) * | 2019-06-28 | 2023-05-26 | 广州华多网络科技有限公司 | Method, device, equipment and storage medium for processing stubble finding picture material |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101493928A (en) * | 2009-02-10 | 2009-07-29 | 国网信息通信有限公司 | Digital watermarking embedding, extracting and quantizing step size coordinating factor optimizing method and device |
CN102129693A (en) * | 2011-03-15 | 2011-07-20 | 清华大学 | Image vision significance calculation method based on color histogram and global contrast |
CN102254295A (en) * | 2011-07-13 | 2011-11-23 | 西安电子科技大学 | Color halftoning image watermarking algorithm based on support vector machine |
CN102509072A (en) * | 2011-10-17 | 2012-06-20 | 上海大学 | Method for detecting salient object in image based on inter-area difference |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2000024312A (en) * | 1998-07-15 | 2000-01-25 | Square Co Ltd | Game apparatus and information recorded medium |
-
2012
- 2012-11-09 CN CN201210447588.6A patent/CN102938139B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101493928A (en) * | 2009-02-10 | 2009-07-29 | 国网信息通信有限公司 | Digital watermarking embedding, extracting and quantizing step size coordinating factor optimizing method and device |
CN102129693A (en) * | 2011-03-15 | 2011-07-20 | 清华大学 | Image vision significance calculation method based on color histogram and global contrast |
CN102254295A (en) * | 2011-07-13 | 2011-11-23 | 西安电子科技大学 | Color halftoning image watermarking algorithm based on support vector machine |
CN102509072A (en) * | 2011-10-17 | 2012-06-20 | 上海大学 | Method for detecting salient object in image based on inter-area difference |
Also Published As
Publication number | Publication date |
---|---|
CN102938139A (en) | 2013-02-20 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US10824910B2 (en) | Image processing method, non-transitory computer readable storage medium and image processing system | |
CN103021002B (en) | Colored sketch image generating method | |
CN104715451B (en) | A kind of image seamless fusion method unanimously optimized based on color and transparency | |
US8879835B2 (en) | Fast adaptive edge-aware matting | |
CA3137297C (en) | Adaptive convolutions in neural networks | |
CN108182671B (en) | Single image defogging method based on sky area identification | |
Tan et al. | Multipoint filtering with local polynomial approximation and range guidance | |
US20210248729A1 (en) | Superpixel merging | |
Zhou et al. | Research on weighted priority of exemplar-based image inpainting | |
Liu et al. | Selective color transferring via ellipsoid color mixture map | |
CN105681686A (en) | Image processing method and system | |
CN102214362B (en) | Block-based quick image mixing method | |
CN108171776B (en) | Method for realizing image editing propagation based on improved convolutional neural network | |
CN102938139B (en) | Automatic synthesis method for fault finding game images | |
Xingjie et al. | The image blending method for face swapping | |
CN104376538A (en) | Image sparse denoising method | |
CN102572305B (en) | Method of video image processing and system | |
CN103413337B (en) | A kind of color fog generation method based on man-machine interactively | |
CN101799931B (en) | Painting rendering method based on colour feature study | |
He et al. | Effective haze removal under mixed domain and retract neighborhood | |
CN102456221B (en) | Method for rapidly eliminating image noise | |
CN105303508B (en) | Image processing method and device | |
Yan et al. | Re-texturing by intrinsic video | |
CN103390080A (en) | Plant disease speckle color simulation method and device | |
Zhao | Interactive texture replacement of cartoon characters based on deep learning model |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant |