Heteroptics analysis device, heteropticallyaltered image generating device for generating heteroptics to create perception approximating original image, heteroptics analysis method, and program
外国特許コード  F150008514 

整理番号  AF1604TW 
掲載日  2015年11月19日 
出願国  台湾 
出願番号  102135306 
公報番号  201419213 
公報番号  I559255 
出願日  平成25年9月30日(2013.9.30) 
公報発行日  平成26年5月16日(2014.5.16) 
公報発行日  平成28年11月21日(2016.11.21) 
優先権データ 

発明の名称 （英語）  Heteroptics analysis device, heteropticallyaltered image generating device for generating heteroptics to create perception approximating original image, heteroptics analysis method, and program 
発明の概要（英語） 
(TWI559255) According to the present invention, in the case of multiresolution decomposition performed on the color components of image data using an orientationselective wavelet frame or an orientationselective filter bank, acquisition of subband signals, and reconstitution of the image by adding together the acquired subband signals, between the decomposition phase of multiresolution decomposition and the synthesis phase, a coefficient process is performed on the decomposition detail coefficients output from the decomposition phase, doing so in such a way that small values are kept smaller the greater the energy of the decomposition detail coefficients in question, while small values are enhanced the lower the energy of the decomposition detail coefficients in question. Alternatively, the coefficient process may be performed in the opposite manner. 
特許請求の範囲（英語） 
[claim1] 1. Kinds of the analysis installments of pseudoblepsia, the department at least has the memory department and control department, this pseudoblepsia's the characteristic of analysis installment is: The above memory department has: The filters memory method, is the memory belongs without the position approximate filters and various plural number positions set of detail filters position selective wavelet frame or the position selective filters group; And the image data memory method, is the memory image data; The above control department has: Decomposes the method, is carries on the multiple resolutions to decompose to obtain the inferior frequency band signal to the above image data color ingredient because of the above position selective wavelet frame or the above position selective filters group; The reconstruction method, because of above inferior frequency band signal adding together that the above decomposition method obtains, takes advantage of this reconstruction phantom to obtain the reconstruction image data; And the pseudoblepsia quantity numerical method, is figures out the above image data and above reconstruction image data ratio or the difference between above color ingredient, takes advantage of this pseudoblepsia quantity the numerical; The above decomposition method department has further: The coefficient processing method, is between the decomposition times and synthesis times that the above multiple resolution decomposes, carries on following coefficient processing to the above decomposition time output's decomposition detail coefficient: This decomposition detail coefficient's energy greatly the small value will suppress slightly, this decomposition detail coefficient's energy is smaller the small value to strengthen. [claim2] 2. Like the analysis installment of application patent scope 1st pseudoblepsia, the above coefficient processing method is carries on individual processing according to above decomposition detail coefficient the difference of plus sign. [claim3] 3. Like the analysis installment of application patent scope 1st pseudoblepsia, the above color ingredient is CIELAB color space L*, a* and b* one. [claim4] 4. Like the analysis installment of application patent scope 3rd pseudoblepsia, the above pseudoblepsia quantity numerical method uses the chromatic aberration, namely uses the L* value and a* value and b* value and above reconstruction image data difference square sum total the square root between the above image data carries on the computation as the above pseudoblepsia quantity. [claim5] 5. Like the analysis installment of application patent scope 3rd pseudoblepsia, the above coefficient processing method is carries on view of above image data a* and/or the b* color ingredient carries on above coefficient processing of following revision: Greatly the small value will suppress from the energy that a* and/or the b* above decomposition detail coefficient and L* above decomposition detail coefficient decides slightly, the above energy will be smaller the small value to strengthen. [claim6] 6. Like the analysis installment of application patent scope 1st pseudoblepsia, the above coefficient processing method is the use, when the above decomposition detail coefficient's energy is big assumes the S character curve automatic continuous change, when the above energy hour assumes the N character curve automatic continuous change the function to carry on above coefficient processing. [claim7] 7. Like the analysis installment of application patent scope 1st pseudoblepsia, the above coefficient processing method is between the above decomposition times and above synthesis times, the standardized above decomposition detail coefficient, the above decomposition detail coefficient after standardization the norm of standardized decomposition detail coefficient as the above energy, to should standardize the decomposition detail coefficient to carry on above coefficient processing, to carries on the above standardized after the coefficient processes the above standardized decomposition detail coefficient again the inverse operation. [claim8] 8. Like the analysis installment of application patent scope 1st pseudoblepsia, the above decomposition method is the use above position or the above position carries on the above multiple resolution decomposition for the multidirection windmill small frame for the horizontal direction, the vertical direction and biorthogonal wavelet filters group of opposite angle direction constitution. [claim9] 9. Like the analysis installment of application patent scope 1st pseudoblepsia, the above decomposition method is in the above multiple resolution decomposition, may use other filters groups, for example according to the exponent number carries on windmill small frame the change of number of times the filters group. [claim10] 10. Like the analysis installment of application patent scope 1st pseudoblepsia, the above multiple resolution that carries on because of the above decomposition method divides the system of solutions to be most greatly redundant the multiple resolution decompositions, to pull out most greatly reduces the multiple resolution decompositions or the parts pulls out reduces the partial repetition multiple resolution decompositions. [claim11] 11. Like the analysis installment of application patent scope 1st pseudoblepsia, the above control department has further the phantom edition method, this phantom edition method is the edition above image data, and carries on the following control: In view of carries on the above multiple resolution decomposition because of above decomposition method after above image data of edition; The above pseudoblepsia quantity numerical method is computation from the above reconstruction image data that passing through the above image data of edition obtains with the above image data that has not edited the ratio or the difference as the above pseudoblepsia quantity; The above phantom edition method is advance makes the above pseudoblepsia quantitative change small relapse edition. [claim12] 12. Like the analysis installment of application patent scope 1st pseudoblepsia, the above control department has further the phantom edition method, this phantom edition method is the edition above image data, and carries on the following control: In view of carries on the above multiple resolution decomposition because of above decomposition method after above image data of edition; The above pseudoblepsia quantity numerical method is computation after above image data of edition with from should among the above reconstruction image data that the above image data of edition obtains the ratio or the difference as the above pseudoblepsia quantity; The above phantom edition method is advance makes the above pseudoblepsia quantity to become the predetermined value relapse edition. [claim13] 13. 1 kinds of pseudoblepsia increase phantoms have installment, the department at least have the memory department and control department, this pseudoblepsia increase phantom has the characteristic of installment is: The above memory department has: The filters memory method, is the memory belongs without the position approximate filters and various plural number positions set of detail filters position selective wavelet frame or the position selective filters group; And the image data memory method, is the memory image data; The above control department has: Decomposes the method, is carries on the multiple resolutions to decompose to obtain the inferior frequency band signal to the above image data because of the above position selective wavelet frame or the above position selective filters group; And the reconstruction method, because of above inferior frequency band signal adding together that the above decomposition method obtains, takes advantage of this reconstruction phantom to obtain the reconstruction image data; The above decomposition method department has further: The coefficient processing method, is between the decomposition times and synthesis times that the above multiple resolution decomposes, carries on following coefficient processing to the above decomposition time output's decomposition detail coefficient: This decomposition detail coefficient's energy greatly the small value will strengthen, this decomposition detail coefficient's energy is smaller the small value to suppress slightly. [claim14] 14. According to claim 13, pseudoblepsia increase phantom has the installment, the above coefficient processing method is carries on individual processing according to above decomposition detail coefficient the difference of plus sign. [claim15] 15. According to claim 13, pseudoblepsia increase phantom has the installment, the above image data's color ingredient is CIELAB color space L*, a* and b* one. [claim16] 16. According to claim 15, pseudoblepsia increase phantom has the installment, the above coefficient processing method is carries on view of above image data a* and/or the b* color ingredient carries on above coefficient processing of following revision: Greatly the small value will strengthen from the energy that a* and/or the b* above decomposition detail coefficient and L* above decomposition detail coefficient decides, the above energy will be smaller the small value to suppress slightly. [claim17] 17. According to claim 13, pseudoblepsia increase phantom has the installment, the above coefficient processing method is the use, when the above decomposition detail coefficient's energy is big assumes the N character curve automatic continuous change, when the above energy hour assumes the S character curve automatic continuous change the function to carry on above coefficient processing. [claim18] 18. According to claim 13, pseudoblepsia increase phantom has the installment, the above coefficient processing method is between the above decomposition times and above synthesis times, the standardized above decomposition detail coefficient, the above decomposition detail coefficient after standardization the norm of standardized decomposition detail coefficient as the above energy, to should standardize the decomposition detail coefficient to carry on above coefficient processing, to carries on the above standardized after the coefficient processes the above standardized decomposition detail coefficient again the inverse operation. [claim19] 19. According to claim 13, pseudoblepsia increase phantom has the installment, the above decomposition method is the use above position for or the above position carries on the above multiple resolution decomposition for the multidirection windmill small frame by the horizontal direction, the vertical direction and biorthogonal wavelet filters group of opposite angle direction constitution. [claim20] 20. According to claim 13, pseudoblepsia increase phantom has the installment, the above decomposition method is in the above multiple resolution decomposition, may use other filters groups, for example according to the exponent number carries on windmill small frame the change of number of times the filters group. [claim21] 21. According to claim 13, pseudoblepsia increase phantom has the installment, the above multiple resolution that carries on because of the above decomposition method divides the system of solutions to be most greatly redundant the multiple resolution decompositions, to pull out most greatly reduces the multiple resolution decompositions or the parts pulls out reduces the partial repetition multiple resolution decompositions. [claim22] 22. 1 pseudoblepsia's analysis methods, are at least are having the pseudoblepsia analysis installment of memory department and in control department carry out, this pseudoblepsia's analysis method's characteristic is: The above memory department has: The filters memory method, is the memory belongs without the position approximate filters and various plural number positions set of detail filters position selective wavelet frame or the position selective filters group; And the image data memory method, is the memory image data; Includes in the following step of above control department execution: Decomposes the step, is carries on the multiple resolutions to decompose to obtain the inferior frequency band signal to the above image data color ingredient because of the above position selective wavelet frame or the above position selective filters group; The reconstruction step, because of above inferior frequency band signal adding together that the above decomposition step obtains, takes advantage of this reconstruction phantom to obtain the reconstruction image data; And the pseudoblepsia quantity numerical step, is figures out the above image data and above reconstruction image data ratio or the difference between above color ingredient, takes advantage of this pseudoblepsia quantity the numerical; The above decomposition step department further includes: The coefficient processing step, is between the decomposition times and synthesis times that the above multiple resolution decomposes, carries on following coefficient processing to the above decomposition time output's decomposition detail coefficient: This decomposition detail coefficient's energy greatly the small value will suppress slightly, this decomposition detail coefficient's energy is smaller the small value to strengthen. [claim23] 23. 1 kinds of pseudoblepsia increase phantoms have method, is at least is having the pseudoblepsia increase phantom of memory department and control department has in the installment to carry out, this pseudoblepsia increase phantom has the method characteristic is: The above memory department has: The filters memory method, is the memory belongs without the position approximate filters and various plural number positions set of detail filters position selective wavelet frame or the position selective filters group; And the image data memory method, is the memory image data; Includes in the following step of above control department execution: Decomposes the step, is carries on the multiple resolutions to decompose to obtain the inferior frequency band signal to the above image data because of the above position selective wavelet frame or the above position selective filters group; And the reconstruction step, because of above inferior frequency band signal adding together that the above decomposition step obtains, takes advantage of this reconstruction phantom to obtain the reconstruction image data; The above decomposition step department further includes: The coefficient processing step, is between the decomposition times and synthesis times that the above multiple resolution decomposes, carries on following coefficient processing to the above decomposition time output's decomposition detail coefficient: This decomposition detail coefficient's energy greatly the small value will strengthen, this decomposition detail coefficient's energy is smaller the small value to suppress slightly. [claim24] 24. 1 kinds of programs, are use to enable at least to have the pseudoblepsia analysis installment of memory department and control department carry out the pseudoblepsia analysis method, this program's characteristic is: The above memory department has: The filters memory method, is the memory belongs without the position approximate filters and various plural number positions set of detail filters position selective wavelet frame or the position selective filters group; And the image data memory method, is the memory image data; Makes carries out the following step at the above control department: Decomposes the step, is carries on the multiple resolutions to decompose to obtain the inferior frequency band signal to the above image data color ingredient because of the above position selective wavelet frame or the above position selective filters group; The reconstruction step, because of above inferior frequency band signal adding together that the above decomposition step obtains, takes advantage of this reconstruction phantom to obtain the reconstruction image data; And the pseudoblepsia quantity numerical step, is figures out the above image data and above reconstruction image data ratio or the difference between above color ingredient, takes advantage of this pseudoblepsia quantity the numerical; Yu Qianshu decomposes in the step to further cause the following step execution: The coefficient processing step, is between the decomposition times and synthesis times that the above multiple resolution decomposes, carries on following coefficient processing to the above decomposition time output's decomposition detail coefficient: This decomposition detail coefficient's energy greatly the small value will suppress slightly, this decomposition detail coefficient's energy is smaller the small value to strengthen. [claim25] 25. 1 kinds of programs, are use to enable at least to have the pseudoblepsia increase phantom of memory department and control department have the installment to carry out the pseudoblepsia increase phantom to have the method, this program's characteristic is: The above memory department has: The filters memory method, is the memory belongs without the position approximate filters and various plural number positions set of detail filters position selective wavelet frame or the position selective filters group; And the image data memory method, is the memory image data; Makes carries out the following step at the above control department: Decomposes the step, is carries on the multiple resolutions to decompose to obtain the inferior frequency band signal to the above image data because of the above position selective wavelet frame or the above position selective filters group; And the reconstruction step, because of above inferior frequency band signal adding together that the above decomposition step obtains, takes advantage of this reconstruction phantom to obtain the reconstruction image data; Yu Qianshu decomposes in the step to further cause the following step execution: The coefficient processing step, is between the decomposition times and synthesis times that the above multiple resolution decomposes, carries on following coefficient processing to the above decomposition time output's decomposition detail coefficient: This decomposition detail coefficient's energy greatly the small value will strengthen, this decomposition detail coefficient's energy is smaller the small value to suppress slightly. 




国際特許分類(IPC) 

参考情報 （研究プロジェクト等）  CREST Alliance for Breakthrough between Mathematics and Sciences (ABMS) AREA 
日本語項目の表示
発明の名称  色知覚の新しい数理モデルによる色知覚の定量化、錯視の分析装置、分析方法、および、プログラム 

※
ライセンスをご希望の方、特許の内容に興味を持たれた方は、問合せボタンを押してください。
『 Heteroptics analysis device, heteropticallyaltered image generating device for generating heteroptics to create perception approximating original image, heteroptics analysis method, and program 』に関するお問合せ
 国立研究開発法人科学技術振興機構（ＪＳＴ） 知的財産マネジメント推進部
 URL: http://www.jst.go.jp/chizai/
 Email:
 Address: 〒1028666 東京都千代田区四番町53
 TEL: 0352148293
 FAX: 0352148476