Vision assistance using mobile telephone

• US 8,331,628 B2
• Filed: 12/08/2009
• Issued: 12/11/2012
• Est. Priority Date: 12/08/2008
• Status: Expired due to Fees

• Current Assignee: Georgios Stylianou, Stavros Papastavrou
• Original Assignee: Georgios Stylianou, Stavros Papastavrou
• Inventors: Stylianou, Georgios, Papastavrou, Stavros
• Primary Examiner(s): Werner, Brian P
• Assistant Examiner(s): GORADIA, SHEFALI DINESH
• Application Number:
  US12/633,288
• Publication Number:
  US 20100144291A1
• Time in Patent Office: 1,099 Days
• Field of Search: None
• US Class Current: 382/114
• CPC Class Codes: G06V 10/421 by analysing segments intersecting the pattern
• CPC Class Codes: G09B 21/006 using audible presentation of the information
• CPC Class Codes: G09B 21/008 using visual presentation of the information for the partially sighted

Abstract

Methods and system for providing vision assistance using a portable telephone with a built-in camera. In some embodiments, the system identifies the value of a bank note by determining the average number of transitions between black and white in each vertical line of pixels corresponding to a numeric digit. In other embodiments, the system captures an image and identifies an object in the image by comparing the value of each pixel in the image to a threshold intensity and marking the pixels that exceed the threshold. The system then generates a plurality of candidate groups by grouping marked pixels that are within a predetermined distance from other marked pixels. The object is identified based on the relative position of each candidate group to other candidate groups.

Claims

1. A method of identifying a value of a bank note using a portable telephone with a built-in camera, the method comprising:

capturing an image of the bank note;

converting the image of the bank note to a black and white image;

locating a numeric value printed on the bank note in the image;

dividing the numeric value into a plurality of vertical lines;

identifying a vertical line that includes only white pixels as an end of a first digit in the numeric value;

determining a number of transitions between black and white in each of a plurality of vertical lines corresponding to the first digit;

calculating an average number of transitions between white and black for the plurality of vertical lines corresponding to the first digit; and

identifying the first digit based on the average number of transitions.

2. The method of claim 1, further comprising

identifying the first digit as a one when the average number of transitions from white to black is one;

identifying the first digit as a zero when the average number of transitions from white to black is two; and

identifying the first digit as one of a two and a five when the average number of transitions from white to black is three.

3. The method of claim 1, further comprising:

identifying a second vertical line that includes only white pixels as an end of a subsequent digit in the numeric value;

determining a number of transitions between white and black in each of a plurality of vertical lines corresponding to the subsequent digit;

calculating the average number of transitions from white to black for the plurality of vertical lines corresponding to the subsequent digit; and

identifying the second digit as a zero when the average number of transitions from white to black is two.

4. The method of claim 1, further comprising identifying the first digit as a two when a number of white pixels in a first vertical line is greater than a number of white pixels in a second vertical line, wherein the first vertical line is a leftmost vertical line corresponding to the first digit and wherein the second vertical line is to the right of the first vertical line.

5. The method of claim 1, further comprising identifying the first digit as a five when a number of white pixels in a first vertical line is less than a number of white pixels in a second vertical line, wherein the first vertical line is a leftmost vertical line corresponding to the first digit and wherein the second vertical line is to the right of the first vertical line.

6. The method of claim 1, wherein locating a numeric value printed on the bank note in the image includes

identifying a first plurality of pixels forming a bottom boundary of the bank note in the image;

identifying a second plurality of pixels forming a left boundary of the bank note in the image;

locating a bottom left corner of the numeric value printed on the bank note in the image by locating a first black pixel that is closest to an intersection of the left boundary and the bottom boundary;

locating a top left corner of the numeric value printed on the bank note by

analyzing a plurality of pixels forming a horizontal line from the bottom left corner of the numeric value,

moving the horizontal line up one pixel when a black pixel is located in the plurality of pixels forming the horizontal line, and

identifying the top left corner of the numeric value as a farthest left pixel in the horizontal line when all pixels in the plurality of pixels forming the horizontal line are white; and

determining a height of the numeric value by calculating a difference between a vertical position of the bottom left corner and a vertical position of the top left corner.

7. The method of claim 6, further comprising estimating a width of the numeric value based on the height of the numeric value.

8. The method of claim 1, further comprising

detecting the presence of multiple numeric digits in the numeric value printed on the bank note;

and

identifying the value of each numeric digit.

9. The method of claim 8, further comprising displaying an error message to a user if two or more numeric digits are identified as non-zero numbers.