Apparatus for reading a license plate

• US 4,817,166 A
• Filed: 05/05/1986
• Issued: 03/28/1989
• Est. Priority Date: 05/05/1986
• Status: Expired due to Term

• Current Assignee: Perceptics LLC
• Original Assignee: Perceptics LLC
• Inventors: Herrera, Juan A., Gonzalez, Rafael C.
• Primary Examiner(s): Boudreau, Leo H.
• Application Number:
  US06/859,639
• Time in Patent Office: 1,058 Days
• Field of Search: 382/1, 382/9, 382/21, 382/22, 382/31, 382/34, 382/51, 382/60, 382/57, 382/48, 340/51, 356/71, 364/478
• US Class Current: 382/105
• CPC Class Codes: G06V 20/625 License plates
• CPC Class Codes: G06V 20/63 Scene text, e.g. street names
• CPC Class Codes: G06V 2201/09 Recognition of logos
• CPC Class Codes: G06V 2201/10 Recognition assisted with metadata
• CPC Class Codes: G08G 1/0175 by photographing vehicles, e.g. when violating traffic rules

Abstract

A video camera produces an image of a license plate on a vehicle, and a scanning apparatus finds a license plate number in the image. The identification of the license plate number is verified in a confidence check section by checking for the presence of a state logo. Next, a character extractor identifies individual characters of the number by finding and tracing a contour along interior portions of the character edges, and the contour length, character height, and character width are then checked in a validity checking section to determine whether they are within predetermined acceptable ranges. To correct for obscuring objects on the license plate, a character alignment section determines the top line and baseline of the license plate number and disregards portions of the characters that appear to be outside of these limits, and a license plate frame correction section is utilized to extract the characters when a portion thereof is obscured by a license plate frame. Once extracted, the characters are recognized by a structural analysis section and a state recognition section recognizes the state logo. Once the state is identified, a verification section rereads the license plate number utilizing knowledge of the type style used by the identified state.

Claims

1. An apparatus for finding and identifying an image of a license plate of a predetermined class that bears characters and that is mounted on a vehicles, comprising:

means for producing an image that includes the license plate;

means for scanning the image and identifying a block of the image that meets predetermined frequency that is above a selected non-zero threshold, said threshold being selected to correspond to the frequency expected to be produced by scanning across the characters of any license plate of the predetermined class; and

decision means for comparing the block to second criteria to determine and indicate whether a block is an image of the license plate characters.

2. The apparatus of claim 1 wherein said decision means comprise:

means for determining the height and width of the block based on the area of the image that meets the predetermined criteria;

comparison means for comparing the height and width of the block to predetermined height and width ranges to decide and indicate whether a block is an image of identifying characters on the license plate.

3. The apparatus of claim 2 further comprising character finding means comprising:

edge finder means for making a first identification of edges of characters within the identified blocks in the image of the license plate, the edges having a range of intensity levels;

edge discriminator means for discriminating between the range of intensity levels of the edges and for selecting and defining a threshold intensity level based on the range of intensity levels of the edges; and

contour tracing means for tracing the contours of characters using the threshold intensity level.

4. The apparatus of claim 3 further comprising:

validity check means for measuring the length of said contours; and

comparison means for comparing the length of said contours to predetermined length criteria and identifying contours that do not meet the predetermined length criteria.

5. The apparatus of claim 2 further comprising character finding means comprising:

background means for determining a background intensity level of the image that corresponds to the most frequently occurring intensity level in the identified block of the image of the license plate;

edge finder means using a gradient operator for making a first identification of edges of potential characters in the image of the license plate;

the edges being defined as areas having an intensity level gradient above a predetermined value and, thus, including a range of edge intensity levels;

edge discriminator means using a Laplacian operator for identifying a selected portion of the edges that have a selected edge intensity level, said selected edge intensity level being within the range of edge intensity levels and being a selected percentage of the range above the background intensity level; and

contour tracing means for tracing the contours of potential characters based on the selected portion of the edges.

6. The apparatus of claim 2 further comprising means for recognizing and extracting characters within the identified block of the license plate.

7. The apparatus of claim 6 further comprising:

validity check means for measuring the individual height and width of the characters on the license plate;

comparison means for comparing the height and width of the characters to predetermined height and width criteria and identifying characters that do not meet the predetermined height and width criteria.

8. The apparatus of claim 2 further comprising:

character alignment means for defining a straight baseline and a straight top line based, respectively, on the lowermost and uppermost points of the characters, whereby portions of potential characters may lay outside of the baseline and top line; and

said alignment means being operable to eliminate portions of the characters that lie partially outside of the baseline and the top line and to define the baseline and the top line as the edges of the characters where character portions lie outside of the baseline and top line.

9. The apparatus of claim 2 further comprising character recognizing means which comprise:

structure analysis means for analyzing the characters to identify the convex hull, bays and holes of each character; and

means for recognizing and identifying the characters based upon the convex hull, bays and holes of a character.

10. An apparatus for finding and identifying an image of a license plate that bears characters and that is mounted on a vehicle, comprising:

means for producing an image that includes the license plate;

means for scanning and image and identifying a block of the image that meet predetermined criteria indicating that the block is possibly an image of the characters of the license plate;

decision means for comparing the block to second criteria to determine and indicate whether the block is an image of the license plate character;

character alignment means for defining a straight baseline and a straight top line based, respectively, on the lowermost and uppermost points of the characters, whereby portions of potential characters may lay outside of the baseline and the top line;

said alignment means being operable to eliminate portions of characters that lie partially outside of the baseline and the top line and to define the baseline and the top line as the edges of the characters where character portions lie outside of the baseline and the top line;

said character alignment means further comprising;

means for identifying the uppermost points and lowermost points on each character;

means for computing a top line that is the best straight line that fits through the uppermost points and for computing a baseline that is the best straight line that fits through the lowermost points;

means for determining a first measure of how well the top line and baseline fit the uppermost and lowermost points, respectively, and comparing said first measure to predetermined criteria of acceptable line fit; and

means for repetitively removing the uppermost and lowermost points of the characters, one point at a time, the most perturb the line fit and for recomputing the top line and baseline until said first measure does meet said predetermined criteria of acceptable line fit.

11. An apparatus for detecting and perceiving a license plate that bears identifying characters and may bear the logo of a state, provide or country, comprising:

means for producing an image including the license plate and for producing and storing digital image data corresponding to the image;

means for scanning the image data and identifying blocks of the image containing the identifying characters;

means for determining the contours of the identifying characters within the blocks;

recognition means for recognizing each character by structural analysis of the contours;

logo finding means for scanning the digital image data corresponding to at least one secondary block of the image within at least one selected area positioned relative to the identifying characters and determining whether said secondary block has an intensity pattern that meets predetermined criteria; and

recognition means for recognizing the state logo by analyzing the digital image data corresponding to the secondary block when said secondary block meets the predetermined criteria.

12. An apparatus for finding an image of a license plate that bears identifying characters and the name of a state, province or country in one of a plurality of expected positions on the plate and that is mounted on a vehicle, comprising:

means for producing an image of the license plate and at least a portion of the vehicle on which the license plate is mounted;

means for scanning the image and identifying at least one block of the image that has intensity transitions that meet a predetermined frequency of occurrence criteria and potentially constitutes an image of the identifying characters on the license plate;

means for determining the height and width of the block of the image that meets the predetermined frequency of occurrence criteria;

decision means for comparing the height and width of the block of the image to predetermined height and width ranges to decide and indicate whether the block is an image of identifying characters on a license plate;

second means for scanning the image in at least one selected area located in a position relative to the block, said selected area including at least one position in which a state, province or country name is expected to occur; and

check means for detecting intensity transitions in the selected area of said second scanning means and for determining and indicating whether the intensity transitions meet second predetermined criteria indicating that the name of a state, province or country is within said selected area, whereby a check is performed to assure that a block of the image is an image of identifying characters of a license plate.

13. An apparatus for reading a character set of a license plate that is partially obscured by a license plate frame, comprising:

means for producing an image including the license plate;

means for scanning the image and identifying at least one block of the image that meets predetermined criteria indicating that the block is possibly a license plate character set;

means for determining the presence of a license plate frame that is partially obscuring the character set;

means for locating the obscuring edge of the license plate frame that is disposed over at least part of the character set; and

means for extracting the individual characters from the character set using the obscuring edge of the license plate frame.

14. The apparatus of claim 13 wherein said locating means traces the contours of the characters and uses the local maxima and minima of the contours to define the obscuring edge of the frame.

15. The apparatus of claim 13 wherein said extracting means defines the characters using the obscuring edge of the frame as edges of the characters.

16. In an apparatus for analyzing characters of an image, the improvement comprising:

background means for determining a background intensity level of the image;

edge finder means for making a first identification of edges of characters on the image, the edges having intensity levels different from the background intensity level and having a range of intensity levels;

edge discriminator means for discriminating between the intensity levels of the edges, for locating a selected portion of the edges having an intensity level range that is smaller than and within the range of edge intensity levels and for defining a threshold intensity level based on the intensity level of the selected portion; and

contour tracing means for tracing the contours of potential characters using the threshold intensity level defined by said edge discriminator means.

17. The improvement of claim 16 wherein said background means is operable to determine a background intensity level of the image that corresponds to the intensity level occurring most frequently in a selected portion of the image.

18. The improvement of claim 16 wherein said edge finder means uses a gradient operator for making a first identification of edges of the characters in the image, the edges being defined as areas having an intensity level gradient above a predetermined value.

19. The improvement of claim 16 wherein said edge discriminator means uses a Laplacian operator for identifying interior portions of the edges.

20. In an apparatus for analyzing characters of an image of a license plate;

the improvement comprising;

means for tracing the contours of the characters in the image of the license plate; and

validity check means for measuring the length of said contours and comparing the length of said contours to predetermined length criteria to reject contours that are not characters.

21. In an apparatus for analyzing characters of an image, the improvement comprising:

means for locating portions of the image that meet predetermined criteria indicating that the portion may be a character and for determining the contour of the portion;

validity check means for measuring the height and width of the contours on the image; and

comparison means for comparing the height and width of the contours to predetermined height and width criteria and reject contours that do not meet the predetermined height and width criteria whereby contours that are not characters are rejected.

22. In an apparatus for analyzing a line of characters on an image, the improvement comprising:

character aligning means for producing a top line and a baseline, said baseline being defined by at least a portion of the lowermost positions of each character and said top line being defined by at least a portion of the uppermost positions on each character, whereby a portion of some of said characters may lie outside the baseline and top line; and

redefining means for redefining the characters to eliminate portions of the characters that lie below the baseline or above the top line.

23. A method for reading a license plate character set made up of individual characters on a license plate on a vehicle comprising:

producing an image including the license plate;

scanning the image and identifying at least one block of the image that meets predetermined criteria indicating that the block is possibly a license plate character set, said predetermined criteria being a pattern of intensity changes occurring at a predetermined frequency that is above a selected non-zero threshold, said threshold being selected to correspond to the frequency expected to be produced by scanning across the characters of any license plate of a predetermined class;

extracting the individual characters from the license plate character set; and

recognizing and identifying each character of the license plate character set.

24. The method of claim 23 wherein said extracting comprises:

identifying the edges of the characters of the license plate number; and

discriminating between the interior portions of the edges and the remainder of the edges, whereby the interior portions of the edges are nearer the character than the remainder; and

defining the interior portions as the contour of the characters.

25. The method of claim 23 wherein said extracting comprises:

detecting the presence of a license plate frame that is partially obscuring the license plate character set; and

disregarding the license plate frame and extracting the characters of the license plate character set.

26. The method of claim 25 wherein said disregarding comprises:

identifying the edge of the license plate frame adjacent to the license plate character set; and

defining the edge of the characters based on the adjacent edge of the frame and distinguishing the characters from the license plate frame.

27. The method of claim 23 wherein said step of extracting comprises:

determining a background intensity level of the image;

finding the edges of the characters;

after the steps of determining a background intensity level and finding the edges, setting a threshold intensity level and based on the background intensity level and the intensity levels of the edges; and

using the threshold intensity level to extract the characters.

28. A method for reading a license plate character set made up of individual characters on a license plate on a vehicle comprising:

producing an image including the license plate;

scanning the image and identifying at least one block of the image that meets predetermined criteria indicating that the block is possibly a license plate character set;

extracting the individual characters from the license plate character set;

recognizing and identifying each character of the license plate character set; and

said extracting further comprising determining the baseline and top line of the license plate character set, disregarding any portion of a character that lies below the baseline and disregarding any portion of a character that lies above the top line.

29. A method for reading the characters of a license plate character set on a license plate on a vehicle, the license plate having a state logo identify a state and having ambiguous characters in the license plate character set, comprising:

producing an image including the license plate;

producing and storing digital image data corresponding to the image;

scanning the image data and identifying at least one block of the image data that meets predetermined criteria indicating that the block is possibly a license plate character set;

extracting the individual characters from the license plate character set;

recognizing and identifying each character of the license plate character set; and

scanning the image data to recognize the state logo and identify the state of the license plate.

30. The method of claim 29 wherein the step of recognizing the state logo is performed after identifying at least one block of the image that meets predetermined criteria.

31. A method for reading the characters of a license plate character set on a license plate on a vehicle, the license plate having a state logo identifying a state and having ambiquous characters in the license plate character set, comprising:

producing an image including the license plate;

scanning the image and identifying at least one block of the image that meets predetermined criteria indicating that the block is possibly a license plate character set;

extracting the individual characters from the license plate character set;

recognizing and identifying each character of the license plate character set;

recognizing the state logo and identifying the state of the license plate;

determining the type style of the identified state'"'"'s license plate; and

analyzing the license plate character set to identify the ambiguous characters in the license plate character set base on the type of the identified state.