# US 2013/0294654 A1 · Robust character segmentation for license plate images

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2013/0294654 A1
- Document ID: US20130294654A1
- Status: Published application
- Kind: A1
- Application number: 13/539,739
- Filed: 2012-07-02
- Published: 2013-11-07
- Earliest claimed: 2012-05-04
- Assignees: Xerox Corporation
- Inventors: Aaron Michael Burry; Claude Fillion; Vladimir Kozitsky
- Web page: https://patentgrep.com/patent/US20130294654A1

## CPC

- G06V 20/62: Text, e.g. of license plates, overlay texts or captions on TV images
- G06V 30/148: Segmentation of character regions
- G06V 20/625: {License plates}
- G06V 30/10: Character recognition
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING

## Abstract

A method and system for achieving accurate segmentation of characters with respect to a license plate image within a tight bounding box image. A vehicle image can be captured by an image capturing unit and processed utilizing an ALPR unit. A vertical projection histogram can be calculated to produce an initial character boundary (cuts) and local statistical information can be employed to split a large cut and insert a missing character. The cut can be classified as a valid and/or a suspect character and the suspect character can be analyzed. The suspect character can be normalized and passed to an OCR module for decoding and generating a confidence quote with every conclusion. The non-character images can be rejected at the OCR level by enforcing a confidence threshold. An adjoining suspect narrow character can be combined and the OCR confidence of the combined character can be assessed.

## Claims

### Claim 1

A license plate character segmentation method, said method comprising:
determining an initial cut utilizing a partitioning;
classifying said cut as a valid character and/or a suspect character in order to analyze said suspect character and enable robust performance over a plurality of state logos, fonts, and character spacing; and
normalizing and passing said suspect character to an optical character recognition subsystem for decoding and generating a confidence quote with every conclusion and thereafter modifying individual character conclusions based on a set of confidence thresholds.

### Claim 2 (depends on 1)

The method of claim 1 wherein said modification of character conclusions comprises rejecting a non-character image by enforcing a confidence threshold.

### Claim 3 (depends on 1)

The method of claim 1 wherein said modification of character conclusions comprises accepting a non-character image by enforcing a confidence threshold.

### Claim 4

The method of claim wherein said modification of character conclusions comprises combining an adjoining suspect narrow character and assessing said optical character recognition confidence of said combined character to achieve enhanced character segmentation accuracy.

### Claim 5

The method of claim wherein said partitioning comprises a vertical projection histogram.

### Claim 6 (depends on 5)

The method of claim 5 further comprising:
computing said vertical projection histogram from a binarized image and determining a midpoint of a segment along said vertical projection histogram having a value at or below said threshold:
utilizing said midpoint as an initial character segmentation boundary and removing an invalid raw block;
determining a location of a maximum histogram value and a beginning and an endpoint for each of said remaining revised block and thereafter measuring a median spacing between said revised block midpoint; and
splitting a wide block and converting a gap between said block into an eligible block and thereafter checking a validity of character and flag if suspect.

### Claim 7 (depends on 5)

The method of claim 5 further comprising:
splitting a single, wide cut into two by placing a boundary at a minimum histogram value in a region near a center of said wide cut; and
replacing a large gap caused by omission of a special character with a missing image data.

### Claim 8 (depends on 1)

The method of claim 1 further comprising training an image-based classifier to identify presence of a valid letter and/or number and applying said image-based classifier to each segmented cuts.

### Claim 9 (depends on 1)

The method of claim 1 further comprising Identifying and addressing at least one segmentation error.

### Claim 10 (depends on 1)

The method of claim 1 wherein said segmentation returns a bitmap of segmented image along with a label identifying said bitmap that are likely a valid character image.

### Claim 11 (depends on 9)

The method of claim 9 wherein identifying said at least one segmentation error further comprises:
receiving a character image and said label and identifying said image utilizing said label as a likely character or not; and
creating a copy of an input character label array as outlabel and initializing all values of said outlabel to zero.

### Claim 12

The method of cairn 11 further comprising:
setting each element in said outlabel to one for a character identified as a non-character that are touching a boundary until a valid character is encountered;
assigning said outlabel element to two for each non-character surrounded by said valid character; and
setting said outlabel element to three for each set of at least two non-characters surrounded by said valid character and combining at least two images and copying to a first location for each set of element in said outlabel set to three and also set outlabel for second location back to one.

### Claim 13 (depends on 9)

The method of claim 9 wherein addressing said at least one segmentation error further comprises:
obtaining an optical character recognition identification and confidence by running each character image labeled as zero, two, and three;
comparing said optical character recognition confidence to said threshold for each image labeled as three, if greater, change in said label from one to three and output said optical character recognition identification as part of said code; and
selecting a lowest confidence neighbor for each image labeled as two, merging two images and passing to said optical character recognition subsystem to obtain an optical character recognition identification and confidence for said merged image.

### Claim 14 (depends on 12)

The method of claim 12 further comprising:
changing said label from one to four and output merged identification as part of said code if confidence of said merged image is greater than said threshold and greater than confidence of said character labeled as two and greater than its neighbor; and
comparing said confidence of said character labeled as two to said threshold if said merged image confidence does not meet a requirement otherwise passing said identification as part of said code.

### Claim 15

A license plate character segmentation system, said system comprising:
a processor;
a data bus coupled to said processor; and
a computer-usable medium embodying computer code, said computer-usable medium being coupled to said data bus, said computer code comprising instructions executable by said processor and configured for:
determining an initial cut utilizing a partitioning;
classifying said cut as a valid character and/or a suspect character in order to analyze said suspect character and enable robust performance over a plurality of state logos, fonts, and character spacing; and
normalizing and passing said suspect character to an optical character recognition subsystem for decoding and generating a confidence quote with every conclusion and thereafter modifying individual character conclusions based on a set of confidence thresholds.

### Claim 16 (depends on 15)

The system of claim 15 wherein said modification of character conclusions comprises rejecting a non-character image by enforcing a confidence threshold.

### Claim 17 (depends on 15)

The system of claim 15 wherein said modification of character conclusions comprises accepting a non-character image by enforcing a confidence threshold.

### Claim 18 (depends on 15)

The system of claim 15 wherein said modification of character conclusions comprises combining an adjoining suspect narrow character and assessing said optical character recognition confidence of said combined character to achieve enhanced character segmentation accuracy.

### Claim 19 (depends on 15)

The system of claim 15 wherein said partitioning comprises a vertical projection histogram.

### Claim 20

A processor-readable medium storing code representing instructions to cause a process to perform a process of license plate character segmentation, said code comprising code to:
determine an initial cut utilizing a partitioning;
classify said cut as a valid character and/or a suspect character in order to analyze said suspect character and enable robust performance over a plurality of state logos, fonts, and character spacing; and
normalize and pass said suspect character to an optical character recognition subsystem for decoding and generating a confidence quote with every conclusion and thereafter modifying individual character conclusions based on a set of confidence thresholds.
