Image pickup system for obtaining flat image without distortion

• US 5,848,197 A
• Filed: 01/06/1997
• Issued: 12/08/1998
• Est. Priority Date: 04/28/1992
• Status: Expired due to Fees

• Current Assignee: Olympus Optical Corporation Limited (Olympus Corporation)
• Original Assignee: Olympus Optical Corporation Limited (Olympus Corporation)
• Inventors: Ebihara, Toshiyuki
• Primary Examiner(s): Razavi, Michael
• Assistant Examiner(s): Patel, Jayanti K.
• Application Number:
  US08/779,058
• Time in Patent Office: 701 Days
• Field of Search: 382/198, 382/204, 382/259, 382/254, 382/156, 382/159, 382/155, 382/300, 382/293
• US Class Current: 382/275
• CPC Class Codes: G06T 5/80 Geometric correction

Abstract

An image reading unit picks up an image to convert the image into an electric signal. An A/D converter converts the electric signal into a digital signal, and an input image memory stores the digital signal. An image reduction processing unit reduces the image data stored in the input image memory. A reduced image memory stores the reduced image. A control point coordinate generating unit reads out the image stored in the reduced image memory to detect distortion. An image converter corrects distortion of the image data loaded from the input image memory on the basis of the information of the detected distortion. The image data whose distortion is corrected is stored in a corrected image memory. The image data whose distortion is corrected and which is stored in the corrected image memory is read out and recorded, so that when an image pickup operation of characters having distortion is to be performed, the distortion is corrected to obtain a flat image, free from distortion, like an image on a flat surface.

Claims

1. An image pickup apparatus comprising:

image pickup means for picking up a target document image to output an electric signal;

converting means for converting the electric signal from said image pickup means into a digital signal to output image data having a plurality of pixels;

image reduction means for substantially reducing a size of an image in the image data by removing some of said plurality of pixels from the image data outputted from said converting means to output reduced image data;

control point coordinates generating means for generating coordinates of at least four control points in accordance with distortion information of the reduced image data from said image reduction means by using a neural network of said control point coordinates generating means which performs a learning operation as to generation of coordinates so as to obtain in advance control point coordinates concerning distortion of arbitrary original image data image data based on said control points being arranged in a matrix pattern relative to the original image data; and

distortion correcting means for correcting distortion of the image data outputted from said converting means by position-converting said image data corresponding to each of the control points coordinates which are generated by said control point coordinates generating means, and by executing interpolation of image data defined by the control points in accordance with three control points from among said at least four control points.

2. An apparatus according to claim 1, wherein said apparatus further comprises:

an input image memory for storing the image data received from said converting means, and

a reduced image memory for storing reduced image data received from said image reduction means.

3. An apparatus according to claim 1, wherein said apparatus further comprises a corrected image memory for storing corrected image data corrected by said distortion correcting means.

4. An apparatus according to claim 1, wherein said neural network performs learning with cross points of vertical and horizontal parallel lines as control points and outputs coordinate data of the control points.

5. An apparatus according to claim 4, wherein said neural network is a hierarchical neural network.

6. An apparatus according to claim 1, wherein said distortion correcting means converts the image data having a position corresponding to the control points, the coordinates of which are generated by said control point coordinates generating means, into image data divided in a lattice manner.

7. An image pickup apparatus comprising:

image pickup means for picking up a target document image to output an electric signal;

converting means for converting the electric signal from said image pickup means into a digital signal to output image data having a plurality of pixels;

image reduction means for substantially reducing a size of an image in the image data by removing some of said plurality of pixels from the image data outputted from said converting means to output reduced image data;

control point coordinates generating means for generating coordinates of at least four control points in accordance with distortion information of the reduced image data from said image reduction means by using a neural network of said control point coordinates generating means which performs a learning operation as to generation of coordinates so as to obtain in advance control point coordinates concerning distortion of arbitrary image data; and

distortion correcting means for correcting distortion of the image data outputted from said converting means by position-converting said image data corresponding to each of the control points coordinates which are generated by said control point coordinates generating means, and by executing interpolation of image data defined by the control points in accordance with three control points from among said at least four control points;

wherein said distortion correcting means executes interpolation of the image data defined by the control points based on the following equations;

space="preserve" listing-type="equation">PiP=s×

PiPj+t×

PiPk

space="preserve" listing-type="equation">QiQ=s×

QiQj+t×

QiQk

where PiP, PiPj, PiPk, QiQ, QiQj and QiQk are vectors, Pi, Pj and Pk represent the coordinates of the three control points before the conversion, Qi, Qj and Qk represent the coordinates of the three control points after the conversion, s is a coefficient of the vectors PiPj and QiQj, and t is a coefficient of the vectors PiPk and QiQk.

8. An apparatus according to claim 7, wherein said apparatus further comprises:

an input image memory for storing the image data received from said converting means, and

a reduced image memory for storing reduced image data received from said image reduction means.

9. An apparatus according to claim 7, wherein said apparatus further comprises a corrected image memory for storing corrected image data corrected by said distortion correction means.

10. An apparatus according to claim 7, wherein said neural network performs learning with cross points of vertical and horizontal parallel lines as control points and outputs coordinate data of the control points.

11. An apparatus according to claim 7, wherein said neural network is a hierarchical neural network.

12. An apparatus according to claim 7, wherein said distortion correcting means converts the image data having a position corresponding to the control points, the coordinates of which are generated by said control point coordinates generating means, into image data divided in a lattice manner.

13. An image pickup apparatus comprising:

image pickup means for picking up a target document image to output an electric signal;

converting means for converting the electric signal from said image pickup means into a digital signal to output image data having a plurality of pixels;

image reduction means for substantially reducing a size of an image in the image data by removing some of said plurality of pixels from the image data outputted from said converting means to output reduced image data;

control point coordinates generating means for generating coordinates of at least four control points in accordance with distortion information of the reduced image data from said image reduction means by using a neural network of said control point is coordinates generating means which performs a learning operation as to generation of coordinates so as to obtain in advance control point coordinates concerning distortion of arbitrary image data; and

distortion correcting means for correcting distortion of the image data outputted from said converting means by position-converting said image data corresponding to each of the control points coordinates which are generated by said control point coordinates generating means, and by executing interpolation of image data defined by the control points in accordance with three control points from among said at least four control points;

wherein said neural network performs learning with cross points of vertical and horizontal parallel lines as control points and outputs coordinate data of the control points; and

wherein said distortion correcting means converts the image data having a position corresponding to the control points, the coordinates of which are generated by said control point coordinates generating means, into image data divided in a lattice manner.