System and method for quantifying reflection e.g. when analyzing laminated documents

1. A system for computerized authentication of a laminated object, the system comprising:

a digital image of at least a portion of the laminated object stored in a computer-implemented memory;

a shininess analyzer operative, using a processor, to generate shininess data quantifying shininess of said digital image; and

a parameterized computerized authentication sub-system operative, using the processor, to differentially perform at least one laminated object authentication operation based on said shininess data,

wherein said image is represented by a multiplicity of pixels and said shininess analyzer is operative for quantifying shininess of the image by:

generating a 2-dimensional histogram array representing frequency of occurrence of each of a plurality of n-tuplets of color component values, in the image,

wherein said shininess analyzer is also operative to identify, for each individual diagonal from among a plurality of diagonals covering the histogram array, a maximum array element corresponding to an n-tuplet of color component values which occurs more frequently than other color component values within the same individual diagonal.

2. A system according to claim 1 wherein said laminated object comprises a physical document.

3. A system according to claim 2 wherein said physical document comprises an instance of a known type of physical ID documents and wherein the system also comprises a data-base storing computerized data characterizing a plurality of known types of physical ID documents.

4. A system according to claim 3 wherein said data characterizing a plurality of known types of physical ID documents comprises data indicating locations of pre-defined fields of interest within each of said plurality of known types of physical ID documents.

5. A system according to claim 1 wherein each said n-tuplet comprises at least a pair of color values from among at least 3 color components representing each of said multiplicity of pixels.

6. A system according to claim 1 wherein said shininess analyzer is also operative for determining a shininess value for each individual pixel, defining an individual n-tuplet of color component values, from among the multiplicity of pixels, by finding said individual n-tuplet of color component values in said 2-dimensional histogram array, thereby to identify a specific diagonal from among the plurality of diagonals to which said individual n-tuplet belongs, and computing a distance between said individual n-tuplet of color component values and the maximum array element defined for said specific diagonal.

7. A system according to claim 6 wherein said shininess analyzer is also operative for outputting a function of said distance.

8. A system according to claim 7 wherein said outputting a function of said distance comprises outputting a binary map including a binarization of said distance for each of said pixels, depending on whether said distance is above or below a predetermined threshold imposed on said distance.

9. A method for quantifying shininess of the image by:

generating a 2-dimensional histogram array representing frequency of occurrence of each of a plurality of n-tuplets of color component values, in the image;

for each individual diagonal from among a plurality of diagonals covering the histogram array, identifying a maximum array element corresponding to an n-tuplet of color component values which occurs more frequently than other color component values within the same individual diagonal;

determining a shininess value for each individual pixel, defining an individual n-tuplet of color component values, from among the multiplicity of pixels, by finding said individual n-tuplet of color component values in said 2-dimensional histogram array, thereby to identify a specific diagonal from among the plurality of diagonals to which said individual n-tuplet belongs, and computing a distance between said individual n-tuplet of color component values and the maximum array element defined for said specific diagonal; and

outputting a function of said distance,

wherein said identifying an n-tuplet which occurs more frequently is operative to disregard at least one frequently occurring color which exceeds a threshold selected to disqualify initially bright colors.

10. A system according to claim 1

wherein said authentication operation comprises authenticating a flow of laminated objects based on information derived from fields of interest of known location within digital images representing said objects, according to at least one pre-defined parameter governing handling of each of said fields, and

wherein said sub-system is operative to differentially perform said authenticating by adjusting said pre-defined parameter for each individual instance within said flow to reflect a lower level of confidence in information derived from an individual field to the extent that the individual field is found by said shininess analyzer to be shiny, and a higher level of confidence in information derived from that individual field, to the extent that the individual field is found by said shininess analyzer not to be shiny.

11. A system according to claim 10 wherein said authentication operation includes at least one forgery test and wherein said parameter is used by a processor to determine whether or not to perform said forgery test based on said individual field, as opposed to performing at least one forgery test based on at least one field other than said individual field.

12. A system according to claim 1 wherein said sub-system is operative to differentially perform at least one authentication operation by determining, based on said shininess data, whether or not to fix said image, by computationally adjusting pixel values to remove estimated shininess effects, before executing a forgery test.

13. A system according to claim 1 wherein said sub-system is operative to differentially perform at least one authentication operation by aggregating information about shininess of various instances of each of at least first and second document types, handling all instances of the first document type using a first forgery test whose parameters are selected given a first level of shininess that appears to characterize said first document type; and handling all instances of the second document type using a second forgery test whose parameters are selected given a second level of shininess that appears to characterize said second document type.

14. A system according to claim 1 wherein said sub-system performs differentially at least one authentication operation by carrying out the authentication operation and reporting its result accompanied by a “level of confidence” indication stipulating that said results enjoy a low level of confidence, only if said shininess data indicates that shininess capable of adversely affecting said results has been detected.

15. A non-transitory computer readable medium having computer readable program code stored therein, said computer readable program code adapted to be executed to implement a method for quantifying shininess of the image by:

generating a 2-dimensional histogram array representing frequency of occurrence of each of a plurality of n-tuplets of color component values, in the image;

for each individual diagonal from among a plurality of diagonals covering the histogram array, using at least one processor for identifying a maximum array element corresponding to an n-tuplet of color component values which occurs more frequently than other color component values within the same individual diagonal;

determining a shininess value for each individual pixel, defining an individual n-tuplet of color component values, from among the multiplicity of pixels, by finding said individual n-tuplet of color component values in said 2-dimensional histogram array, thereby to identify a specific diagonal from among the plurality of diagonals to which said individual n-tuplet belongs, and computing a distance between said individual n-tuplet of color component values and the maximum array element defined for said specific diagonal; and

outputting a function of said distance,

wherein said identifying an n-tuplet which occurs more frequently is operative to disregard at least one frequently occurring color lying along an individual diagonal whose distance from an initial color along the individual diagonal indicates that said frequently occurring color is such a dark color that the initial color could not be a shiny version of that dark color.