VALUABLE DOCUMENT IDENTIFICATION METHOD AND SYSTEM

Abstract

A valuable document identification method and system are provided. The method comprises, detecting features in different space ranges of a valuable document, and obtaining multi-source information Ω={Xi, Xj, . . . , Xn}, wherein Xi∩Xj≠φ, or Xi∩Xj≠φ and Xi<=>Xj; labeling the space position of Xj with Xi according to semantic constraints of Xi and Xj, and obtaining position constraints Ψij(x, y); extracting a characteristic value fi from Xi, and extracting a characteristic value fj from Xj according to the position constrains Ψij(x, y); determining whether fi, fj meet the characteristic criteria of the valuable document, if yes, then receiving the valuable document, or else, rejecting the valuable document. The method enables improved reliability and robustness of the valuable document identification system.

Description

The present application claims the priority of Chinese Patent Application No. 201110130431.6, titled “Method and system for Identifying Valuable Document”, filed with the Chinese State Intellectual Property Office (CSIPO) on May 19, 2011, the entire disclosure of which are incorporated herein by reference.

FIELD OF THE INVENTION

The present invention relates to the field of valuable document identifying technique, and in particular to a method and system for identifying valuable document based on multi-source information cooperation.

BACKGROUND OF THE INVENTION

In FIG. 1, a schematic structural diagram of an existing system for identifying valuable document based on result integration is shown. The system for identifying valuable document collects characteristics of the valuable document by a sensor 1, a sensor 2 and a sensor 3 to obtain a characteristic 1, a characteristic 2 and a characteristic 3. Then characteristics are identified respectively to obtain identification results, and the identification results are integrated together according to a “logic AND” policy. An identification result indicating that “the valuable document is legal” is output if the identification result of each characteristic is legal; else an identification result indicating that “the valuable document is illegal” is output.

The existing system for identifying valuable document has following disadvantages. Although multiple sensors are employed to collect various characteristics of the valuable document, context relation or constraint among individual characteristics is unconsidered during the identifying process, a separate process is performed on each of the characteristics respectively, and the identification results of characteristics are integrated together simply according to the “logic AND” policy when the result is outputted. Different characteristics can not be validated and checked mutually, the reliability of the system is poor; moreover, no context constraint is established among characteristics, so that the final judging result can only be obtained by judge and calculate all the characteristics, causing calculation redundancy.

SUMMARY OF THE INVENTION

An embodiment of the present invention provides a method and system for identifying valuable document to improve reliability and robustness of the system for identifying valuable document.

A method for identifying valuable document provided by an embodiment of the present invention includes:

S1. Detecting characteristics of the valuable document in different spatial scales to obtain multi-source information Ω={X_i, X_j, . . . , X_n};

wherein X_i indicates the characteristic in the i-th space scale, X_j indicates the characteristic in the j-th space scale, 1≦i≦n≦, 1≦j≦n, and i≠j; semantic constraints on X_i and X_j are: X_i∩X_j≠φ;

S2. Determining a spatial location of X_j by using X_i to obtain a location constraint Ψ_ij(x,y) according to the semantic constraints on X_i and X_j;

S3. Extracting a feature value f_i from the X_i; and extracting a feature value f_j from the X_j according to the location constraint Ψ_ij(x,y), checking whether f_i and f_j conform to the semantic constraint on X_i and X_j, judging that f_i and f_j are legal if yes; else outputting a rejection identification,

wherein f_i is an optimal characteristic of the i-th spatial scale, and f_j is an optimal characteristic of the j-th spatial scale;

S4. Judging whether f_i and f_j conform to a characteristic standard of the valuable document; outputting the reception identification to receive the valuable document if yes; else outputting the rejection identification to reject the valuable document.

A system for identifying valuable document provided by an embodiment of the present invention includes:

a multi-source information detecting module adapted for detecting characteristics of the valuable document in different spatial scales to obtain multi-source information {X_i, X_j, . . . , X_n}; where X_i indicates the characteristic in the i-th space scale, X_j indicates the characteristic in the j-th space scale, 1≦i≦n, 1≦j≦n, and i≠j; semantic constraints on X_i and X_j are: X_i∩X_j≠φ;

a cooperative sensing module adapted for determining a spatial location of X_j by using X_i according to the semantic constraints on X_i and X_j to obtain a location constraint Ψ_ij(x,y); extracting a feature value f_i from X_i; and extracting a feature value f_j from X_j according to the location constraint Ψ_ij(x,y), checking whether f_i and f_j conform to the semantic constraint on X_i and X_j, determining that f_i and f_j are legal if yes; else outputting a rejection identification, wherein f_i is an optimal characteristics of the i-th spatial scale, and f_j is an optimal characteristics of the j-th spatial scale; and

a cooperative decision module adapted for judging whether f_i and f_j conform to a characteristic standard of the valuable document; outputting a reception identification to receive the valuable document if yes; else outputting a rejection identification to reject the valuable document.

By implementing embodiments of the present invention, the following advantages can be obtained:

The method and the system for identifying valuable document provided by the embodiment of the present invention obtain the multi-source information and establish the context scene of the multi-source information by detecting characteristics of the valuable documents in different spatial scales; the collaborative process is performed on the multi-source information of the valuable document by using the context constraint of the multi-source information during the identifying process, and thus the applicability of the system is expanded, the information of the detected object is obtained with higher accuracy and reliability, the consistent explanation and description of the detected object is established more accurately, the reliability and the robustness of the system for identifying valuable document is enhanced, and the calculation efficiency of the system is improved.

BRIEF DESCRIPTION OF THE DRAWINGS

The invention will be further understood with reference to drawings which are only a part of the specification and explain the invention in conjunction with the embodiment of the present invention. The invention is not limited by the drawings. In the drawings:

FIG. 1 is a schematic structural diagram of a system for identifying valuable document based on result integration in prior art;

FIG. 2 is a schematic flow chart of a method for identifying valuable document provided by a first embodiment of the present invention;

FIG. 3 is a schematic diagram of a process for obtaining multi-source information provided by an embodiment of the present invention;

FIG. 4 is a schematic diagram of a semantic constraint on multi-source information of a valuable document provided by an embodiment of the present invention;

FIG. 5 is a schematic flow chart of a cooperative sensing portion in a method for to identifying valuable document provided by a second embodiment of the present invention;

FIG. 6 is a schematic flow chart of a cooperative decision portion in a method for identifying valuable document provided by a second embodiment of the present invention; and

FIG. 7 is a schematic structural diagram of a system for identifying valuable document provided by a third embodiment of the present invention.

DETAILED DESCRIPTION OF THE INVENTION

The technical solution in the embodiment of the present invention will be described clearly and completely in conjunction with the drawings in the embodiment of the present invention hereinafter. Obviously, the described embodiments are only a part but not all of the embodiments of the present invention. All the other embodiments obtained by those skilled in the art without creative work based on the embodiment in the present invention pertain to the scope of protection of the present invention.

Reference is made to FIG. 2, which is a schematic flow chart of a method for identifying valuable document provided by a first embodiment of the present invention, and the method includes the following steps:

S1. Detecting characteristics of the valuable document in different spatial scales to obtain multi-source information Ω={X_i, X_j, . . . , X_n}, where X_i indicates the characteristic in the i-th space scale, X_j indicates the characteristic in the j-th space scale, 1≦i≦n, 1≦j≦n, and i≠j; semantic constraints on X_i and X_j are:

X_i∩X_j≠φ, or X_i∩X_j≠φ and X_i<=>X_j;

S2. Determining a spatial location of X_j with X_i according to the semantic constraints on X_i and X_j to obtain a location constraint Ψ_ij(x,y);

S3. Extracting a feature value f_i from the X_i; and extracting a feature value f_j from the X_j according to the location constraint Ψ_ij(x,y);

S4. Judging whether f_i and f_j conform to a characteristic standard of the valuable document; receiving the valuable document if yes; else rejecting the valuable document.

In the above step S1, as required by practical application, the characteristics of the valuable document in different spatial scales can be detected by a specific array of sensors to obtain multi-source information. The multi-source information includes spectral information, magnetic information, material information and other physical information.

The characteristic of the valuable document has the feature of “global cooperation and local competition”, the characteristics in different spaces not only keep relatively independent, but also keep a certain semantic constraint in the context (scene) for describing the valuable document: redundancy, complementation or correlation. Individual characteristics work cooperatively in the context circumstance for describing the valuable document, consisting integral description of the valuable document. As shown in FIG. 3, the characteristics of the valuable document in different spatial scales is detected by a sensor 1, a sensor 2, a sensor 3, . . . , and a sensor n to obtain respectively the characteristics X₁, X₂, X₃, . . . , and X_n. There exists redundancies between X₁ and X₂, and between X₃ and X_n, there exists correlation between X₁ and X_n, and they cooperatively constitute the consistent description or explanation of the valuable document. This procedure obtains the information of the valuable document in different spatial scales by building the array of sensors, for providing multi-source information for the following cooperative sensing process.

The above steps S2 and S3 are the cooperative sensing processes. Spatial characteristics of the valuable document keep relatively independent locally, which have integrity for describing the nature thereof; and keep “cooperative” relationship globally and conform to a certain semantic constraint. By establishing the context scene of individual spatial characteristics, the optimal characteristics of individual spaces are extracted according to the idea of “local competition and global cooperation”, considering completely the context constraint on individual characteristics, during the identification process. The optimal characteristic refers to the feature value that is legal and most adequate for representing the nature of this space. In the above step S3, the feature value f_i extracted from X_i and the feature value f_j extracted from X_j are both optimal characteristic.

The above step S4 is the cooperative decision process, the decision sorting is performed based on practical application scene according to the optimal characteristics of individual spaces obtained during the cooperative sensing process. Locally, when one feature value is determined as an illegal input, the match of the other spatial characteristics is stopped, a rejection identification is output to reject to identify the valuable document; and globally, only when all the feature values are determined as legal, the identification result is output.

The method for identifying valuable document provided by a second embodiment of the present invention will be described in detail only by taking the multi-source information Ω including spectral information X₁, magnetic information X₂ and material information X₃ as an example in conjunction with FIGS. 4 to 6 hereinafter.

In this embodiment, the multi-source information of the valuable document is defined as: spectral information X₁, magnetic information X₂ and material information X₃. Specifically, the 2-dimensional spectral information X₁ is collected by a spectral sensor, the 1-dimensional magnetic information X₂ is collected by a magnetic sensor, and the 1-dimensional material information X₃ is collected by a material sensor. The 2-dimensional spectral information X₁ forms the image of the valuable document, the 1-dimensional magnetic information X₂ records the magnetic signal information which is collected during the valuable document passing, and the 1-dimensional material information X₃ records the thickness information which is collected during the valuable document passing.

1. The Semantic Constraint on the Spectral Information X₁ and the Magnetic Information X₂

For the valuable document, the spectral information X₁ is used for representing the image information of the valuable document, the magnetic information X₂ is used for representing the collected magnetic signal information of the magnetic carrier (e.g., magnetic security thread on the valuable document). The spectral information X₁ contains the optical imaging information of the magnetic carrier, the location “coordinate” of the magnetic carrier can be accurately determined by the spectral information X₁. Therefore, the magnetic sensor forms the semantic constraint on the spectral information X₁ and the magnetic information X₂ for the collecting point “coordinate” of the magnetic carrier on the valuable document, which can be specifically described by the following formula:

X ₁ ∩X ₂≠φ and X₁<=>X₂;  (1)

In FIG. 4, a schematic diagram of the semantic constraint on the multi-source information is shown, in which the black belt-shaped region (i.e., the region B) is the optical image of the magnetic security thread; and the inclined stripe region (i.e., the region A and the region C) is the optical image of the other portion in the valuable document. The magnetic information X₂ represents the magnetic signal information of the magnetic security thread, the optical imaging of the magnetic security thread is contained in the optical imaging X₁ of the whole valuable document, then X₁∩X₂≠φ and X₁<=>X₂. Because the optical reflectivity of the magnetic security thread is low, the optical image thereof is the black belt-shaped region in FIG. 4, and the gray value thereof is significantly lower than that of the ambient region. The location of the magnetic security thread in the image can be determined by analyzing the variance in the gray value of the image. This is the inherent relation between the spectral information and the magnetic security thread (magnetic information) of the valuable document.

2. The Semantic Constraint on the Spectral Information X₁ and the Material Information X₃

The material information X₃ is used for representing the material thickness of the valuable document, and the spectral information X₁ contains the optical imaging information of various material thickness varying regions of the valuable document. The variance in the material thickness of the valuable document reflects onto the strength of the transmitted or reflected spectral energy, which is the inherent instinct relation between the spectral information and the material information. Therefore, the variance in the material thickness (i.e., the material information X₃) of the valuable document can be reflected by the spectral information X₁. The semantic constraint on the spectral information X₁ and the material information X₃ can be described by the following formula:

X ₁ ∩X ₃≠φ and X₁<=>X₃;  (2)

As shown in FIG. 4, the optical imaging lightness value of the magnetic security thread, i.e., the region B, is low, and the optical imaging lightness value of the other regions, such as the regions A and C, is high. From the region A to the region B, the material of the valuable document varies, its corresponding optical imaging lightness value experiences the variation from “high” to “low”; and from the region B to the region C, the material of the valuable document varies, its corresponding optical imaging lightness value experiences the variation from “low” to “high”. Therefore, the part of specific material on the valuable document can be accurately determined by the spectral information X₁.

3. The Semantic Constraint on the Magnetic Information X₂ and the Material Information X₃

The magnetic region in the valuable document has particular material characteristic, and the particularity of this region can be reflected by the material information X₃. The material information X₃ contains the thickness information of the magnetic carrier, and the semantic constraint on the magnetic information X₂ and the material information X₃ can be described by the following formula:

X ₂ ∩X ₃≠φ;  (3)

As shown in FIG. 4, the thickness of the magnetic security thread (i.e., the region B) is much thicker than that of the ambient material. From the region A to the region B, their corresponding thickness values experience the variation from “low” to “high”; and from the region B to the region C, their corresponding thickness values experience the variation from “high” to “low”.

Reference is made to FIG. 5, which is a schematic flow chart of a cooperative sensing portion of a method for identifying valuable document provided by a second embodiment of the present invention. According to the above three semantic constraints, i.e., the semantic constraints (1), (2) and (3), the cooperative sensing procedure of the valuable document is as follow.

S21. Determining a location of the magnetic carrier in the image according to the variety state of the image gray value of the spectral information X₁, and obtaining a magnetic information location constraint Ψ₁₂(x,y);

S22. Determining locations of various materials in the image according to the variety state of the transmittance or refractive index in the spectral information X₁, and obtaining a material information location constraint Ψ₁₃(x,y);

S31. Extracting a feature value f₁ from the spectral information X₁ based on a main component analyzing method;

S32. Extracting a feature value f₂ from the magnetic information X₂ based on the magnetic information location constraint Ψ₁₂(x,y);

Specifically, the magnetic information location constraint Ψ₁₂(x,y) determines the location of the magnetic carrier (e.g. the magnetic security thread) in the valuable document. The feature value f₂ of the magnetic carrier can be accurately extracted from the magnetic information X₂ based on the Ψ₁₂(x,y). The feature value f₂ is the time sequence of the magnetic information.

S33. Extracting a feature value f₃ from the material information X₃ according to the material information location constraint Ψ₁₃(x,y);

Specifically, the material information location constraint Ψ₁₃(x,y) determines the locations of various materials in the valuable document, and the feature value f₃ of a certain material can be accurately extracted from the material information X₃ based on the Ψ₁₃(x,y). For example, the material information X₃ is the thickness information, and the thickness of the magnetic security thread or the thickness of the paper material (such as the paper material other than the magnetic security thread) can be accurately extracted from the material information X₃ based on the Ψ₁₃(x,y).

S34. Checking whether f₂ and f₃ conform to the semantic constraint on X₂ and X₃, and determining that f₂ and f₃ are legal if yes;

Because the production of the valuable document (such as banknote) complies with a strict technology standard, the spectral characteristic and the magnetic characteristic of the real banknote strictly comply with the standard value, without exceeding the range of the standard value. For example, the magnetic security thread of the banknote has specific magnetic information and thickness value. In this embodiment, according to the semantic constraint on X₂ and X₃, the feature values f₂ and f₃ is checked mutually to verify the legality of f₂ and f₃.

S35. In the case that f₂ and f₃ are legal, checking whether an correlation attribute among f₁, f₂ and f₃ conforms to the valuable document standard; and determining that f₁ is legal, if yes; and

S36. Outputting f₁, f₂ and f₃, if f₁, f₂ and f₃ are all legal; else outputting a rejection identification.

Moreover, in the above flow of cooperative sensing valuable document, the location information of the magnetic information X₂ can also be determined by the material information X₃ to obtain the magnetic information location constraint Ψ₃₂(x,y). After the feature values f₁, f₂ and f₃ are extracted based on the main component analyzing method, the legality of f₂ and f₃ can be verified by Ψ₁₂(x,y), Ψ₁₃(x,y), Ψ₃₂(x,y), rather than by verifying whether f₂ and f₃ conform to the semantic constraint on X₂ and X₃. That is to say, the variable f₂ can be substituted into Ψ₁₂(x,y), and f₂ is determined as legal when Ψ₁₂(x,y) is satisfied. In the same way, the variable f₃ can be substituted into Ψ₁₃(x,y) to check whether f₃ is legal; and variables f₂ and f₃ are substituted into Ψ₃₂(x,y) to check the legality of f₂ and f₃.

The cooperative decision process in a method for identifying valuable document provided by a second embodiment of the present invention is as follow:

S41. Judging whether f₁ conforms to a real banknote spectral information data standard; performing S42 if yes, else performing S45;

S42. Judging whether f₂ conforms to a real banknote magnetic information data standard; performing S43 if yes, else performing S45;

S43. Judging whether f₃ conforms to a real banknote material information data standard; performing S44 if yes, else performing S45;

S44. Outputting a reception identification; and

S44. Outputting a rejection identification.

In the system for identifying valuable document, the valuable document that is transferred into the detection region is easily subject to abnormity, such as incline, misplacement, and fold. For example, as shown in FIG. 4, assuming that the detecting direction of the magnetic sensor is the y-axis direction, the location of the magnetic security thread is offset when the valuable document is inclined in the detecting region, and thus only a part of the magnetic information, even no magnetic information, can be collected by the magnetic sensor. Further, the material sensor for collecting the thickness of the magnetic security thread is fixedly mounted in the specific location, and thus the material information of the area other than the magnetic security thread will be collected by the material sensor if the valuable document in the detecting region is offset. In this way, if the identification is performed directly by the information collected by the sensor, rejecting to identify or false identifying can be caused in the system, and thus the reliability is lowered.

According to the method for identifying valuable document provided by the embodiment of the present invention, after the spectral information X₁, the magnetic information X₂ and the material information X₃ of the valuable document are obtained, the feature value f₂ of the magnetic carrier is extracted from the magnetic information X₂ according to the magnetic information location constraint Ψ₁₂(x,y); and the feature value f₃ of a certain material is extracted from the material information X₃ according to the material information location constraint Ψ₁₃(x,y). Even if the valuable document in the detecting region is easily subject to abnormity, such as incline, misplacement, fold, the optimal characteristics of the valuable document in different spatial scales can be extracted accurately, and the reliability of the system for identifying valuable document can be enhanced.

In FIG. 6, a schematic flow chart of a more specific embodiment of the cooperative decision process is shown.

In this embodiment, the decision set Φ₁(), Φ₂() and Φ₃() is employed for indicating decision rules of individual identification sub-tasks, which are specifically as follows.

Φ₁()—spectral information reality deciding rule which indicates the match with the real banknote spectral information data standard. If Φ₁()<T₁, the current valuable document is rejected; and if Φ₁()≧T₁, the current valuable document is received, in which T₁ is the preset threshold range of the spectral information.

Φ₂()—magnetic information reality deciding rule which indicates the match with the real banknote magnetic information data standard. If Φ₂()<T₂, the current valuable document is rejected; and if Φ₂()≧T₂, the current valuable document is received, in which T₂ is the preset threshold range of the magnetic information.

Φ₃()—material information reality deciding rule which indicates the match with the real banknote material information data standard; if Φ₃()<T₃, the current valuable document is rejected; and if Φ₃()≧T₃, the current valuable document is received, in which T₃, is the preset threshold range of the material information.

If the identification results of individual sub-tasks satisfy the cooperative rule Ξ(Φ₁, Φ₂, Φ₃), the “reception” identification is output; else the “rejection” identification is output. The cooperative rule is represented by the following formula:

$\begin{array}{cc}\Xi \left(•\right)=\{\begin{array}{cc}1& {\Phi }_1\ge T_1,{\Phi }_2\ge T_2,{\Phi }_3\ge T_3\\ 0& \mathrm{other}\end{array}& \left(4\right)\end{array}$

As shown in FIG. 6, the cooperative decision procedure is as follow.

I. Rejecting if no legal characteristic is extracted during the cooperative sensing process;

II. Substituting the individual spatial optimal feature values f₁, f₂, and f₃ into the cooperative decision rule Ξ(Φ₁, Φ₂, Φ₃), if the feature values f₁, f₂, and f₃ are extracted during the cooperative sensing process;

III. Stopping calculating other attributes to avoid an invalid calculation if any of Φ₁(f₁)<T₁, Φ₂(f₂)<T₂, and Φ₃(f₃)<T₃ is satisfied, outputting the “rejection” identification, and turning to step V;

IV. Substituting Φ_i(f_i)(i=1, 2, 3) into the cooperative decision rule Ξ; outputting the “reception” identification if yes; else outputting the “rejection” identification; and

V. Ending the calculation.

In the method for identifying valuable document provided by the present embodiment, by performing the collaborative decision, according to the local competition principle, different technical characteristics having semantic constraints provide mutually parameters for determining. When any one of the technical characteristics is determined as illegal, the calculation of the other attribute is stopped to avoid the invalid calculation, so as to improve the calculating efficiency. According to the global cooperative principle, mutual determination is performed on different characteristics to improve the reliability and accuracy of the system for identifying.

It is to be understood by those skilled in the art that all or a part of the processes in the above example method can be implemented by instructing related hardware with computer program, which can be stored in a computer readable storage medium, and the program when performed can include the procedure of the example of the above individual methods. The storage medium can be magnetic disk, optical disk, Read-Only Memory (ROM), Random Access Memory (RAM), or the like.

Correspondingly, the present invention further provides a system for identifying valuable document to achieve all the steps of the method for identifying valuable document in the above embodiment.

Reference is made to FIG. 7, which is a schematic structural diagram of an embodiment of a system for identifying valuable document provided by the present invention.

The system for identifying valuable document according to this embodiment includes:

a multi-source information detecting module adapted for detecting characteristics of the valuable document in different spatial scales to obtain multi-source information Ω={X_i, X_j, . . . , X_n}; where X_i indicates the characteristic in the i-th space scale, X_j indicates the characteristic in the j-th space scale, 1≦i≦n, 1≦j≦n, and i≠j; semantic constraints on X_i and X_j are: X_i∩X_j≠φ; or X_i∩X_j≠φ and X_i<=>X_j;

a cooperative sensing module adapted for determining a spatial location where X_j is located with X_i according to the semantic constraints on X_i and X_j, and obtaining a location constraint Ψ_ij(x,y); extracting a feature value f_i from the X_i; and extracting a feature value f_j from the X_j according to the location constraint Ψ_ij(x,y); and

a cooperative decision module adapted for judging whether f_i and f_j conform to a characteristic standard of the valuable document; receiving the valuable document if yes; else rejecting the valuable document.

Specifically, the multi-source information detecting module includes:

a spectral information detecting device adapted for obtaining spectral information X₁of the valuable document;

a magnetic information detecting device adapted for obtaining magnetic information X₂ of the valuable document; and

a material information detecting device adapted for obtaining material information X₃ of the valuable document;

The spectral information X₁ is used for representing image information of the valuable document;

The magnetic information X₂ is used for representing magnetic signal information of a magnetic carrier of the valuable document, the spectral information X₁ contains optical imaging information of the magnetic carrier, and semantic constraints on X₁ and X₂ are: X₁∩X₂≠φ and X₁<=>X₂;

The material information X₃ is used for representing material thickness of the valuable document, the spectral information X₁ contains optical imaging information of various material thickness varying regions of the valuable document, and semantic constraints on X₁ and X₃ are: X₁∩X₃≠φ and X₁<=>X₃; and

The material information X₃ contains thickness information of the magnetic carrier, and a semantic constraint on X₂ and X₃ is: X₂∩X₃≠φ.

Specifically, the collaborative sensing module includes:

a magnetic information location constraint processing unit adapted for determining a location of the magnetic carrier in the image according to the variety state of the image gray value of the spectral information X₁ to obtain a magnetic information location constraint Ψ₁₂(x,y); and the magnetic information location constraint processing unit also adapted for determining a location of magnetic information X₂ according to the material information X₃ to obtain magnetic information location constraint Ψ₃₂(x,y);

a material information location constraint processing unit adapted for determining locations of various materials in the image according to the variety state of the transmittance or refractive index in the spectral information X₁, and obtaining a material information location constraint Ψ₁₃(x,y);

• • a spectral characteristic extracting unit adapted for extracting a feature value f₁ from the spectral information X₁ based on a main component analyzing method;
  • a magnetic characteristic extracting unit adapted for extracting a feature value f₂ from the magnetic information X₂ based on the magnetic information location constraint Ψ₁₂(x,y);
  • a material characteristic extracting unit adapted for extracting a feature value f₃ from the material information X₃ according to the material information location constraint Ψ₁₃(x,y); and
  • a feature value legality judging unit adapted for checking whether f₂ and f₃ conform to the semantic constraint on X₂ and X₃, and if yes, determining that f₂ and f₃ are legal and checking whether an associate attribute among f₁, f₂ and f₃ conform to the valuable document standard; if the associate attribute among f₁, f₂ and f₃ conform to the valuable document standard, determining that f₁ is legal and outputting f₁, f₂ and f₃; else outputting an rejection identification.

Specifically, the cooperative decision module includes:

• • a spectral information realness judging unit adapted for judging whether f₁ conforms to a real banknote spectral information data standard;

a magnetic information realness judging unit adapted for judging whether f₂ conforms to a real banknote magnetic information data standard;

• • a material information realness judging unit adapted for judging whether f₃ conforms to a real banknote material information data standard; and
  • a synthetic decision unit adapted for outputting an identification of receiving the valuable document if f₁, f₂, and f₃ all conform to the standard; else outputting an identification of rejecting the valuable document.

According to the method and the system for identifying valuable document provided by the embodiment of the present invention, the multi-source information is obtained and the context scene of the multi-source information is established by detecting characteristics of the valuable documents in different spatial scales; the cooperative process is performed on the multi-source information of the valuable document by using the context constraint on the multi-source information during the identifying process, the covering range of the system is expanded, the information of the detected object is obtained with higher accuracy and reliability, the consistent explanation and description of the detected object is established more accurately, the reliability and the robustness of the system for identifying valuable document is enhanced, and the calculating efficiency of the system is improved.

The above are the preferred embodiments of the present invention, it is to be noted that, several modifications and retouches can be made by those skilled in the art without deviating from the principle of the present invention, which are also seemed as within the scope of protection of the present invention.

Claims

1. A method for identifying valuable document, comprising: S1. detecting characteristics of the valuable document in different spatial scales to obtain multi-source information Ω={Xi, Xj, . . . , Xn},wherein Xi indicates the characteristic in the i-th space scale, Xj indicates the characteristic in j-th space scale, 1≦i≦n, 1≦j≦n, and i≠j; semantic constraints on Xi and Xj are: Xi∩Xj≠φ;S2. determining a spatial location of Xj by using Xi to obtain a location constraint Ψij(x,y) according to the semantic constraints on Xi and Xj;S3. extracting a feature value fi from the Xi; and extracting a feature value fj from the Xj according to the location constraint Ψij(x,y), checking whether fi and fj conform to the semantic constraint on Xi and Xj, determining that fi and fj are legal if yes; else outputting a rejection identificationwherein fi is an optimal characteristic of the i-th spatial scale, and fj is an optimal characteristic of the j-th spatial scale; andS4. judging whether fi and fj conform to a characteristic standard of the valuable document; outputting the reception identification to receive the valuable document if yes; else outputting the rejection identification to reject the valuable document.

2. The method for identifying valuable document according to claim 1, wherein the multi-source information comprises spectral information X1, magnetic information X2 and material information X3, wherein the spectral information X1 is adapted for representing image information of the valuable document;the magnetic information X2 is adapted for representing magnetic signal information of a magnetic carrier of the valuable document, the spectral information X1 contains optical imaging information of the magnetic carrier, and semantic constraints on X1 and X2 are: X1∩X2≠φ;the material information X3 is adapted for representing material thickness of the valuable document, the spectral information X1 contains optical imaging information of various material thickness variety regions of the valuable document, and semantic constraints on X1 and X3 are: or X1∩X3≠φ; andthe material information X3 contains thickness information of the magnetic carrier, and a semantic constraint on X2 and X3 is: X2∩X3≠φ.

3. The method for identifying valuable document according to claim 1, wherein the semantic constraint on Xi and Xj are Xi∩Xj≠φ and Xi<=>Xj.

4. The method for identifying valuable document according to claim 3, wherein the multi-source information comprises spectral information X1, magnetic information X2 and material information X3; the spectral information X1 is adapted for representing image information of the valuable document;the magnetic information X2 is adapted for representing magnetic signal information of a magnetic carrier of the valuable document, the spectral information X1 contains optical imaging information of the magnetic carrier, and semantic constraints on X1 and X2 are: X1∩X2≠φ and X1<=>X2;the material information X3 is adapted for representing material thickness of the valuable document, the spectral information X1 contains optical imaging information of various material thickness variety regions of the valuable document, and semantic constraints on X1 and X3 are: X1∩X3≠φ and X1<=>X3; andthe material information X3 contains thickness information of the magnetic carrier, and a semantic constraint on X2 and X3 is: X2∩X3≠φ.

5. The method for identifying valuable document according to claim 1, wherein the step S2 comprises: determining a location of the magnetic carrier in the image according to the variety state of the image gray value of the spectral information X1, and obtaining a magnetic information location constraint Ψ12(x,y); anddetermining locations of various materials in the image according to the variety state of the transmittance or refractive index in the spectral information X1, and obtaining a material information location constraint Ψ13(x,y).

6. The method for identifying valuable document according to claim 5, wherein the step 2 further comprises: determining the location of magnetic information according to the material information X3, and obtaining a magnetic information location constraint Ψ32(x,y).

7. The method for identifying valuable document according to claim 1, wherein the step S3 comprises: S31. extracting a feature value f1 from the spectral information X1 based on a main component analyzing method;S32. extracting a feature value f2 from the magnetic information X2 according to the magnetic information location constraint Ψ12(x,y) based on the main component analyzing method;S33. extracting a feature value f3 from the material information X3 according to the material information location constraint Ψ13(x,y) based on the main component analyzing method;S34. checking whether f2 and f3 conform to the semantic constraint on X2 and X3, and determining that f2 and f3 are legal if yes;S35. in the case that f2 and f3 are legal, checking whether an associate attribute among f1, f2 and f3 conform to the valuable document standard; and if yes, determining that f1 is legal; andS36. outputting f1, f2 and f3, if f1, f2 and f3 are all legal; else outputting a rejection identification.

8. The method for identifying valuable document according to claim 7, wherein the valuable document is real banknote and the step S4 comprises: S41. judging whether f1 conforms to a real banknote spectral information data standard; performing S42 if yes, else performing S45;S42. judging whether f2 conforms to a real banknote magnetic information data standard; performing S43 if yes, else performing S45;S43. judging whether f3 conforms to a real banknote material information data standard; performing S44 if yes, else performing S45;S44. outputting a reception identification; andS44. outputting a rejection identification.

9. A system for identifying valuable document, comprising: a multi-source information detecting module adapted for detecting characteristics of the valuable document in different spatial scales to obtain multi-source information Ω={Xi, Xj, . . . , Xn}; wherein Xi indicates the characteristic in the i-th space scale, Xj indicates the characteristic in the j-th space scale, 1≦i≦n, 1≦j≦n, and i≠j; semantic constraints on Xi and Xj are: Xi∩Xj≠φ;a cooperative sensing module adapted for determining a spatial location of Xj by using Xi according to the semantic constraints on Xi and Xj to obtain a location constraint Ψij(x,y); extracting a feature value fi from Xi; and extracting a feature value fj from the Xj according to the location constraint Ψij(x,y), checking whether fi and fj conform to the semantic constraint on Xi and Xj, determining that fi and fj are legal if yes; else outputting a rejection identification, wherein fi is an optimal characteristics of the i-th spatial scale, and fj is an optimal characteristics of the j-th spatial scale; anda cooperative decision module adapted for judging whether fi and fj conform to a characteristic standard of the valuable document; outputting a reception identification to receive the valuable document if yes; else outputting a rejection identification to reject the valuable document.

10. The system for identifying valuable document according to claim 9, wherein the multi-source information detecting module comprises: a spectral information detecting device adapted for obtaining spectral information X1 of the valuable document;a magnetic information detecting device adapted for obtaining magnetic information X2 of the valuable document; anda material information detecting device adapted for obtaining material information X3of the valuable document, whereinthe spectral information X1 is adapted for representing the image information of the valuable document;the magnetic information X2 is adapted for representing magnetic signal information of a magnetic carrier of the valuable document, the spectral information X1 contains optical imaging information of the magnetic carrier, and semantic constraints on X1 and X2 are: X1∩X2≠φ;the material information X3 is adapted for representing material thickness of the valuable document, the spectral information X1 contains optical imaging information of various material thickness varying regions of the valuable document, and semantic constraints on X1 and X3 are: X1∩X3≠φ; andthe material information X3 contains thickness information of the magnetic carrier, and a semantic constraint on X2 and X3 is: X2∩X3≠φ.

11. The system for identifying valuable document according to claim 9, wherein the multi-source information detecting module adapted for defining the semantic constraints on Xi and Xj are: Xi∩Xj≠φ, and Xi<=>Xj after obtaining the multi-source information Ω={Xi, Xj, . . . , Xn}; thenthe semantic constraints on the spectral information X1 obtained by the spectral information detecting device and the magnetic information X2 obtained by the magnetic information detecting device are: X1∩X2≠φ and X1<=>X2; andthe semantic constraints on the spectral information X1 obtained by the spectral information detecting device and the material information X3 obtained by the material information detecting device are: X1∩X3≠φ and X1<=>X3.

12. The system for identifying valuable document according to claim 9, wherein the cooperative sensing module comprises: a magnetic information location constraint processing unit adapted for determining a location of the magnetic carrier in the image according to the variety state of the image gray value of the spectral information X1, and obtaining a magnetic information location constraint Ψ12(x,y); anda material information location constraint processing unit adapted for determining locations of various materials in the image according to the variety state of the transmittance or refractive index in the spectral information X1, and obtaining a material information location constraint Ψ13(x,y).

13. The system for identifying valuable document according to claim 12, wherein the magnetic information location constraint processing unit adapted for determining the location of magnetic information according to the material information X3, and obtaining a magnetic information location constraint Ψ32(x,y).

14. The system for identifying valuable document according to claim 9, wherein the cooperative sensing module further comprises: a spectral characteristic extracting unit adapted for extracting a feature value f1 from the spectral information X1 based on a main component analyzing method;a magnetic characteristic extracting unit adapted for extracting a feature value f2 from the magnetic information X2 according to the magnetic information location constraint Ψ12(x,y) based on the main component analyzing method;a material characteristic extracting unit adapted for extracting a feature value f3 from the material information X3 according to the material information location constraint Ψ13(x,y) based on the main component analyzing method; anda feature value legality judging unit adapted for checking whether f2 and f3 conform to the semantic constraint on X2 and X3; if yes, determining that f2 and f3 are legal and checking whether an associate attribute among f1, f2 and f3 conform to the valuable document standard; if the associate attribute among f1, f2 and f3 conform to the valuable document standard, determining that f1 is legal and outputting f1, f2 and f3; else outputting the rejection identification.

15. The system for identifying valuable document according to claim 14, wherein the valuable document is real banknote, and the cooperative decision module comprises: a spectral information realness judging unit adapted for judging whether f1 conforms to a real banknote spectral information data standard;a magnetic information realness judging unit adapted for judging whether f2 conforms to a real banknote magnetic information data standard;a material information realness judging unit adapted for judging whether f3 conforms to a real banknote material information data standard; anda synthetic decision unit adapted for outputting the reception identification if f1, f2, and f3 all conform to the standard; else outputting a rejection identification.