US 6,188,801 B1Granted

Method and apparatus for automatic image calibration for an optical scanner

OwnerJenn-Tsair Tsai, Bill Cheninventors

Life of this patent

Estimating the term…

Loading patent…

Abstract

A memory is divided into several blocks in advance for storing calibrated pixels in sequence. Then, start to check each pixel row by row. The obliquity ratio of each pixel will be computed based on two referential blocks on the two sides of the calibration paper. The obliquity ratio will then divided by the length of one pixel. If the residue of the division is larger than half the length of the pixel, it indicates that that pixel is the beginning of next block. Following this rule, we can divide each row into several blocks. The skew pixels will be calibrated according to a first calibration procedure and a second calibration procedure. The calibrated pixels will be stored in the memory blocks in sequence according to their block numbers. After all the pixels of the same row have been calibrated, the calibrated pixels will be read from the memory blocks in sequence. The calibrated procedure repeats until each row of the image has been calibrated.

Claims12

Dependent claims

A method for automatic image calibration comprising the steps of:

a. initializing an index i for X axis and j for Y axis, and a block number;
b. reading a pixel on As (i,j), wherein said As (i,j) representing coordinates for a skew image;
c. computing an obliquity ratio of said pixel on As (i,j);
d. determining if said obliquity ratio divided by the length of said pixel on As (i,j) has a residue less than half the length of said pixel on As (i,j);
e. when said residue is smaller than half the length of said pixel on As (i,j), executing first calibration procedure to generate a calibrated pixel for said pixel on As (i,j), wherein said first calibration procedure follows the formula of (As(i,j)*((b2−a1−a4)/b2))+(As(i,j−1)*(a2/b2))+(As(i+1,j)*((a3/b2)), wherein said b representing the length of said pixel on As (i,j) , a1 representing left calibration portion of said pixel on As (i,j), a2 representing top calibration portion of said pixel on As (i,j), a3 representing right calibration portion of said pixel on As (i,j), a4 representing bottom calibration portion of said pixel on As (i,j);
f. when said residue is larger than half the length of said pixel on As (i,j), executing second calibration procedure to generate a calibrated pixel for said pixel on As (i,j) and incrementing said block number by one, wherein said second calibration procedure follows the formula of (As(i,j)*((a4/b2))+(As(i,j+1)*(a3/b2))+(A(i,j−1)*((a2/b2));
g. storing said calibrated pixel in a memory according to said block number;
h. increasing said index i and said index j by one;
and
i. repeating said step (b) and said step (h) until said index i is equal to a predetermined number and reading said calibrated pixels from said memory according said block number.

7 more elements

A method for automatic image calibration comprising steps of:

a. initializing an index i for X axis and j for Y axis, and a block number;
b. reading a pixel on As (i, j), wherein said As (i,j) representing coordinates for a skew image;
c. computing an obliquity ratio of said pixel on as (i,j), wherein said obliquity ratio is determined by l tan α, l representing the length of said pixel on As(i,j), α representing the angle of said pixel on As(i,j);
d. determining if said obliquity ratio divided by the length of said pixel on As (i,j) has a residue less than half the length of said pixel on As (i,j);
e. when said residue is smaller than half the length of said pixel on As (i,j), executing first calibration procedure to generate a calibrated pixel for said pixel on As (i,j);
f. when said residue is larger than half the length of said pixel on As (i,j), executing second calibration procedure to generate a calibrated pixel for said pixel on As (i,j) and incrementing said block number by one;
g. storing said calibrated pixel in a memory according to said block number;
h. increasing said index i and said index j by one;
and
i. repeating said step (b) and said step (h) until said index i is equal to a predetermined number and reading said calibrated pixels from said memory according said block number.

7 more elements

Dependencies are inferred from the claim text.