Recent Patents in Signal Processing (November 2016) – Handwriting recognition

For our November 2016 issue, we cover recent patents granted in the area of handwriting recognition, proposing novel hardware and software tools and signal processing algorithms for character separation and recognition, word extraction and content analysis.

Patent no 9,465,985 presents methods, systems, and computer-readable media related to a technique for providing handwriting input functionality on a user device. A handwriting recognition module is trained to have a repertoire comprising multiple non-overlapping scripts and capable of recognizing tens of thousands of characters using a single handwriting recognition model. The handwriting input module provides real-time, stroke-order and stroke-direction independent handwriting recognition for multi-character handwriting input. In particular, real-time, stroke-order and stroke-direction independent handwriting recognition is provided for multi-character, or sentence level Chinese handwriting recognition. User interfaces for providing the handwriting input functionality are also disclosed.

Patent no. 9,384,403 presents a system and method that is able to recognize a user's natural superimposed handwriting without any explicit separation between characters. The system and method is able to process single-stroke and multi-stroke characters. It can also process cursive handwriting. Further, the method and system can determine the boundaries of input words either by the use of a specific user input gesture or by detecting the word boundaries based on language characteristics and properties. The system and method analyzes the handwriting input through the processes of fragmentation, segmentation, character recognition, and language modeling. At least some of these processes occur concurrently through the use of dynamic programming.

A method described in patent no 9,349,045 operates an input apparatus for an electronic appliance in a motor vehicle, by having a sensing device sense a character trace from a user while the latter draws the character trace in order to input a character or a group of characters on an input panel of the sensing device. A handwriting recognition device performs character recognition for the sensed character trace, and a display device produces a display for a result of the recognition. To allow comprehensible appliance operation by handwriting recognition in a motor vehicle, the display device also displays the character trace following the performance of the character recognition.

Following the invention no. 9,336,190, a stand alone low cost writing pad includes a rechargeable battery, a low capacity memory, a low power processor, a first pair of connectors and supports audio, video and digital ink capturing functionalities. The writing pad may be detached from and re-attached to a stand alone base unit using the first pair of connectors. The base unit includes another rechargeable battery, high capacity memory, high power processor, and a second pair of connectors. The base unit receives captured audio and digital ink from the writing pad via the communication pathway and the high power processor runs voice recognition and optical character recognition software on received data to generate second data. The second data is displayed on the writing pad and/or stored in the high capacity memory for future use.

A handwriting input recognizing method from patent no. 9,317,144 comprises following steps. Recognize an initial touch point of a touch. Determine an input region associated with the initial touch point. Display the input region. Recognize and display a handwritten input in the input region. Determine whether there is a handwritten input reaches the boundary of the input region. If yes, recognize and display the letters associated with the handwritten input and determine whether the input region receives another handwritten input during a predefined time duration. If yes, determine the sting of letters associated with the handwritten input is a first portion of a complete word, and the sting of letters associated with the another handwritten input is a last portion of the complete word. Combine the first portion and the last portion into a complete word. display letters of the first portion and the last portion in a substantially same size.

Methods and systems for recognizing Devanagari script handwriting are provided in patent no. 9,251,412. A method may include receiving a handwritten input and determining that the handwritten input comprises a shirorekha stroke based on one or more shirorekha detection criteria. Shirorekha detection criteria may be at least one criterion such as a length of the shirorekha stroke, a horizontality of the shirorekha stroke, a straightness of the shirorekha stroke, a position in time at which the shirorekha stroke is made in relation to one or more other strokes in the handwritten input, and the like. Next, one or more recognized characters may be provided corresponding to the handwritten input.

In Patent no. 9,170,734 a handwriting recognition system  is disclosed with a plurality of input modes, including a storage unit, a touch screen display unit, and a processing unit. The storage unit stores character fonts; each of the character fonts corresponds to at least one of the input modes. The touch screen display unit includes an input panel having at least two input areas located thereon; each of the areas inputs data in one of the input modes. The processing unit judges an input set of touch signals to recognize a character input by handwriting in one of the input modes, according to a position on which one of the input areas, which position corresponds to at least one of the set of touch signals. The character is represented by the set of touch signals. A handwriting recognition method applicable to an electronic apparatus with a plurality of input modes is also provided.

Patent no. 8,995,770 deals with performing word recognition operations to determine what an image of a word represents. The method includes accessing a first image. The first image represents an image version of a word. The method further includes accessing a second image. The second image also represents an image version of a word. Using a warp mesh, the method includes warping the second image to cause the second image to approximately match the first image by applying a mesh to the second image and moving vertices of the mesh to warp the second image. The difference between the warped second image and the unwarped first image are determined.

A first technique of recognizing content is disclosed in patent no. 8,989,492, including: determining a first value representative of a pixel content present at a first set of pixels associated with a first distance from a pixel under consideration; determining a second value representative of a pixel content present at a second set of pixels associated with a second distance from the pixel under consideration; and using the first and second values to compute one or more spatial features associated with the pixel under consideration for purposes of content recognition. A second technique of recognizing content is also disclosed, including: determining, for a pixel, a first value representative of a first feature associated with a set of pixels associated with a first direction from the pixel; and determining, for the pixel, a second value representative of a second feature associated with a set of pixels associated with a second direction from the pixel.

If you have an interesting patent to share when we next feature patents related to handwriting recognition, or if you are especially interested in a signal processing research field that you would like to highlight in this section, please send email to Csaba Benedek (benedek.csaba AT sztaki DOT mta DOT hu).

References

Number: 9,465,985
Title:  Managing real-time handwriting recognition
Inventors:  Xia; Mei-Qun (San Francisco, CA), Dolfing; Jannes G. (Daly City, CA), Dixon; Ryan S. (Mountain View, CA), Groethe; Karl M. (San Francisco, CA), Misra; Karan (Mountain View, CA), Bellegarda; Jerome R. (Saratoga, CA), Meier; Ueli (Santa Cruz, CA)
Issued: October 11, 2016
Assignee: Apple Inc. (Cupertino, CA)

Number: 9,384,403
Title: System and method for superimposed handwriting recognition technology
Inventors:  Wimmer; Zsolt (Nantes, FR), Perraud; Freddy (Nantes, FR), Lallican; Pierre-Michel (Nantes, FR), Aradilla; Guillermo (Nantes, FR)
Issued: July 5, 2016
Assignee: MyScript (Nantes, FR)

Number: 9,349,045
Title: Motor vehicle having an input device for handwriting recognition
Inventors: Bouaziz; Tahar (Ingolstadt, DE), Betz; Michael (Ingolstadt, DE)
Issued: May 24, 2016
Assignee: AUDI AG (Ingolstadt, DE)

Number: 9,336,190
Title:  Writing pad with synchronized background audio and video and handwriting recognition
Inventors: Bennett; James D. (Hroznetin, CZ)
Issued:May 10, 2016
Assignee: ENPULZ, LLC (Austin, TX)

Number: 9,317,144
Title: Electronic device, handwriting input recognition system, and method for recognizing handwritten input thereof
Inventors: Mao; Hai-Jun (Shenzhen, CN), Liu; Te-Jia (Shenzhen, CN), Liang; Hai-Sen (Shenzhen, CN), Cheng; Hua-Dong (Shenzhen, CN), Chiang; Chih-San (New Taipei, TW)
Issued: April 19, 2016
Assignee: Fu Tai Hua Industry (Shenzhen) Co., Ltd. (Shenzhen, CN), Hon Hai Precision Industry CO., LTD. (New Taipei, TW)

Number: 9,251,412
Title: Segmentation of devanagari-script handwriting for recognition
Inventors:  Keysers; Daniel Martin (Stallikon, CH), Deselaers; Thomas (Zurich, CH), Rowley; Henry Allan (Sunnyvale, CA)
Issued: February 2, 2016
Assignee: Google Inc. (Mountain View, CA)

Number: 9,170,734
Title: Multiple-input handwriting recognition system and measure thereof
Inventors: Lee; Chia Chung (Shanghai, CN)
Issued:October 27, 2015
Assignee: Inventec Appliances (Pudong) Corporation (Shanghai, CN), Inventec Appliances Corp. (New Taipei, TW), Inventec Appliances (Shanghai) Co., Ltd. (Shanghai, CN)

Number: 8,995,770
Title: Word warping for offline handwriting recognition
Inventors: Kennard; Douglas J. (Provo, UT), Barrett; William Arthur (Provo, UT), Sederberg; Thomas Warren (Orem, UT)
Issued: March 31, 2015
Assignee: Brigham Young University (Provo, UT)

Number: 8,989,492
Title: Multi-resolution spatial feature extraction for automatic handwriting recognition
Inventors: Bellegarda; Jerome R. (Saratoga, CA), Dolfing; Jannes G. A. (Sunnyvale, CA)
Issued: March 24, 2015
Assignee: Apple Inc. (Cupertino, CA)