EP0096219A2
Expired / Lapsed
EP
⏱️ 0.0 years remaining
Image classification method based on histogram analysis
Assignee: International Business Machines Corp
•
Filing: 1983-05-02
•
Granted: —
•
Expires: 2003-05-02
Abstract
Abstract A method is described for classifying a document as bilevel or continuous gray scale for the purpose of selecting a compression algorithm for efficient transmission and/or storage of the digitized image of that document. After the document has been at least partially scanned with a gray scale video camera and digitized pel values have been stored, a pel distribution table or histogram is formed. The histogram is analyzed to determine whether there is a peak region having at least 50% of the total number of pels. If there is not, the document is classified as continuous gray scale. If such a region is found, tail (T) and tail factor (TF) values are computed. The relative magnitudes of these values and the algebraic sign of the T value are used to classify the scanned document as continuous gray scale, white on black or black on white. Once the document is classified, an optimum data compression algorithm can be selected.
Inventors
- 👤 Charles Anthony Kunzinger
- 👤 Dana Alfred Lasher
- 👤 William Karl Steltzenmuller
Bibliographic Details
Application Number
—
Jurisdiction
EP
Data Source
google_patents
Last Synchronized
2026-10-27 19:43
CPC Classifications
H; H04; H04N; H04N1/00; H04N1/40; H04N1/40062
Claims Analysis
Total Claims: 7
Claim 1
Independent Claim
136 words
A method of automatically classifying a document as nominally bilevel or as continuous gray scale for use an image processing system having means for scanning nominally bilevel or continuous gray scale documents, and means for digitizing and storing the captured data wherein the data is stored as pels, each of which has one of a predetermined number of gray scale values, said method being characterized by the steps of : generating a histogram reflecting the distribution of pel values for a given number of pels constituting at least a representative sample of the captured data; analyzing the histogram to determine whether it contains a pel peak region meeting predetermined criteria; and classifying the scanned document as nominally bilevel or continuous gray scale as a function of the magnitude and location of the identified pel peak region.
Claim 2
Independent Claim
136 words
A method of automatically classifying a document as nominally bilevel or as continuous gray scale for use an image processing system having means for scanning nominally bilevel or continuous gray scale documents, and means for digitizing and storing the captured data wherein the data is stored as pels, each of which has one of a predetermined number of gray scale values, said method being characterized by the steps of : generating a histogram reflecting the distribution of pel values for a given number of pels constituting at least a representative sample of the captured data; analyzing the histogram to determine whether it contains a pel peak region meeting predetermined criteria; and classifying the scanned document as nominally bilevel or continuous gray scale as a function of the magnitude and location of the identified pel peak region.
Claim 3
Dependent Claim
31 words
A method as defined in Claim 1 characterized in that it includes the additional step of selecting one of a set of compression algorithms as a function of the image classification.
Claim 4
Dependent Claim
48 words
A method as defined in Claim 1 characterized in that adjacent pel values are grouped into zones, each of said zones encompassing a predetermined number of successive pel values and having a magnitude equal to the total number of pels having any one of the encompassed pel values.
Claim 5
Dependent Claim
100 words
A method as defined in Claim 3 characterized in that the step of analyzing the histogram to determine whether it contains a pel peak region meeting predetermined criteria comprises the further steps of : locating the zone having the greatest number of pels; summing the pel count for at least one zone on each side of the located zone with the pel count for the located zone to establish a peak region pel count; and determining whether the peak region pel count is equal to or greater than a predetermined percentage of the total number of pels in the histogram.
Claim 6
Dependent Claim
30 words
A method as defined in Claim 4 characterized in that a gray scale data compression algorithm is selected when the peak region pel count is less than the predetermined percentage.
Claim 7
Dependent Claim
138 words
A method as defined in Claim 4 characterized in that it includes the following steps when the peak region pel count exceeds the predetermined percentage : computing a tail value T and a tail factor value TF in accordance with the equations where H is the highest zone number containing any pels, L is the lowest zone number containing any pels, PEAK is the zone number having the greatest number of pelsand K1 and K2 are constants; selecting a gray scale algorithm when the absolute magnitude of T is less than the magnitude of TF; determining the algebraic sign of T when the absolute magnitude of T is greater than or equal to TF and selecting either an appropriate black on white algorithm when T is positive or an appropriate white on black algorithm when T is negative.
References Cited (Backward Citations)
No citation records indexed.
USPTO Annuity & Maintenance Windows
Maintenance schedule is calculated for granted US utility patents.