WO2012075821A1 - 一种重叠书写的手写输入方法 - Google Patents

一种重叠书写的手写输入方法 Download PDF

Info

Publication number
WO2012075821A1
WO2012075821A1 PCT/CN2011/077950 CN2011077950W WO2012075821A1 WO 2012075821 A1 WO2012075821 A1 WO 2012075821A1 CN 2011077950 W CN2011077950 W CN 2011077950W WO 2012075821 A1 WO2012075821 A1 WO 2012075821A1
Authority
WO
WIPO (PCT)
Prior art keywords
stroke
character
recognition
input
color
Prior art date
Application number
PCT/CN2011/077950
Other languages
English (en)
French (fr)
Inventor
镇立新
龙腾
Original Assignee
上海合合信息科技发展有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 上海合合信息科技发展有限公司 filed Critical 上海合合信息科技发展有限公司
Priority to EP11846641.6A priority Critical patent/EP2650767A4/en
Priority to KR1020137014847A priority patent/KR101825154B1/ko
Priority to US13/992,392 priority patent/US9052755B2/en
Priority to JP2013542348A priority patent/JP5784141B2/ja
Publication of WO2012075821A1 publication Critical patent/WO2012075821A1/zh

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
    • G06F3/01—Input arrangements or combined input and output arrangements for interaction between user and computer
    • G06F3/018—Input/output arrangements for oriental characters
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
    • G06F3/01—Input arrangements or combined input and output arrangements for interaction between user and computer
    • G06F3/03—Arrangements for converting the position or the displacement of a member into a coded form
    • G06F3/033—Pointing devices displaced or positioned by the user, e.g. mice, trackballs, pens or joysticks; Accessories therefor
    • G06F3/0354—Pointing devices displaced or positioned by the user, e.g. mice, trackballs, pens or joysticks; Accessories therefor with detection of 2D relative movements between the device, or an operating part thereof, and a plane or surface, e.g. 2D mice, trackballs, pens or pucks
    • G06F3/03545—Pens or stylus
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
    • G06F3/01—Input arrangements or combined input and output arrangements for interaction between user and computer
    • G06F3/048—Interaction techniques based on graphical user interfaces [GUI]
    • G06F3/0487—Interaction techniques based on graphical user interfaces [GUI] using specific features provided by the input device, e.g. functions controlled by the rotation of a mouse with dual sensing arrangements, or of the nature of the input device, e.g. tap gestures based on pressure sensed by a digitiser
    • G06F3/0488—Interaction techniques based on graphical user interfaces [GUI] using specific features provided by the input device, e.g. functions controlled by the rotation of a mouse with dual sensing arrangements, or of the nature of the input device, e.g. tap gestures based on pressure sensed by a digitiser using a touch-screen or digitiser, e.g. input of commands through traced gestures
    • G06F3/04883—Interaction techniques based on graphical user interfaces [GUI] using specific features provided by the input device, e.g. functions controlled by the rotation of a mouse with dual sensing arrangements, or of the nature of the input device, e.g. tap gestures based on pressure sensed by a digitiser using a touch-screen or digitiser, e.g. input of commands through traced gestures for inputting data by handwriting, e.g. gesture or text
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/70—Determining position or orientation of objects or cameras
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
    • G06V30/10—Character recognition
    • G06V30/14—Image acquisition
    • G06V30/142—Image acquisition using hand-held instruments; Constructional details of the instruments
    • G06V30/1423—Image acquisition using hand-held instruments; Constructional details of the instruments the instrument generating sequences of position coordinates corresponding to handwriting
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20076—Probabilistic image processing

Definitions

  • the invention belongs to the technical field of text input, and relates to a handwriting input method, in particular to a handwriting input method for overlapping writing.
  • the handwriting input method generally refers to the user writing a character by using a writing pen or a finger on the touch screen of the electronic device, and the electronic device recognizes the written track of the written text, converts it into a corresponding text and displays it on the screen.
  • the user's written track is generally displayed on the display screen in real time, so that the user can see the stroke track feedback information of the written text in time.
  • the writing method is generally to sense the movement track of a human finger on a capacitive touch screen, due to the size limitation of the touch screen. And the resolution limit of the finger contact, the user can only write one or two characters on the touch screen at a time, and each time the user writes a character, it usually pauses for a period of time (typical residence time is about 200ms ⁇ 1000ms), and then waits After the mobile device detects and recognizes the character written last time, the handwriting input of the next character can be performed, so that the input of the character by handwriting is inefficient, and continuous handwriting input cannot be performed according to the habit of natural writing.
  • this method is still limited by the size of the touch screen, and it is impossible to perform continuous writing of any multi-character (for example, writing a complete sentence), especially on mobile electronic devices (such as smart phones) with limited touch screen size, usually
  • the number of written characters that can be accommodated in the full screen of the touch screen is generally small (for example, an iPhone smartphone, usually a screen with a finger
  • the number of characters that can be written is less than 4), and if the user makes a glue or overlap between different characters when writing in full screen, the recognition engine cannot correctly segment and recognize the characters.
  • the present invention utilizes an overlapping writing method to perform a handwritten text input method for arbitrary multi-word continuous writing, which can effectively solve the above problems.
  • the so-called overlapping writing method refers to a writing method in which the user writes a character to completely overlap or partially overlap the first character after writing a character.
  • the method can make the user write different characters without stopping, and can continuously write a plurality of handwritten characters on the handwriting screen and perform corresponding recognition processing, thereby greatly improving the efficiency of handwritten text input.
  • the technical problem to be solved by the present invention is: providing a handwriting input method for overlapping writing, by determining the initial stroke of the overlapping characters, and assigning different display colors to the strokes of the currently written overlapping characters and the last completed character strokes, Thereby, the display is beautiful and clear and does not interfere with the user's handwriting input.
  • the present invention uses the following technical solutions:
  • a handwriting input method for overlapping writing includes the following steps: Step S111: A stylus or a finger touches the touch screen, and starts to input a stroke;
  • Step S112 The stylus or the finger moves on the touch screen, records and displays the stroke of the stroke on the handwritten area on the touch screen, and draws the current stroke of the stroke with the given color A;
  • Step S11 3 The stylus or finger leaves the touch screen, and the input of the current stroke ends;
  • Step S114 determining whether the currently written stroke and all the strokes of the color A that have been input before belong to the same character; if the answer is yes, go to step S124; if the answer is no, go to the step
  • Step S115 determining whether there is a character of the inactive view layer on the screen; the character of the inactive view layer refers to the character that has been completely written and recognized and color-converted last time, and the character is given in step S117. If the answer is yes, go to step S116; if the answer is no, then go to step S117;
  • Step S116 Clearing a character image of the inactive view layer and related stroke information
  • Step S117 Defining a character composed of all strokes having a color value A other than the current stroke is an inactive view layer character
  • Step S118 Submit the current inactive view layer character data to the recognition engine for identification, and output the recognition result;
  • Step S119 determining whether the recognition result given by the recognition engine is sufficiently reliable; if the answer is yes, then proceeding to step S120; otherwise, proceeding to step S122;
  • Step S120 gradually fade the stroke color of the currently inactive view layer character until it disappears and does not display
  • Step S121 Clear the character image of the inactive view layer and the related stroke information; Go to step S123;
  • Step S122 Convert the stroke color of the currently inactive view layer character to another color B for display;
  • Step S123 determining whether there is still a new stroke input, if the answer is yes, then proceeding to step S112; otherwise, proceeding to step S126;
  • Step S124 Keep the current stroke color unchanged
  • Step S125 determine whether there is still a new stroke input, if the answer is yes, then proceeds to step S112, otherwise proceeds to step S126;
  • Step S126 submit the currently written character data to the recognition engine for identification, and output the recognition result
  • step S114 it is determined whether the currently written stroke and the previously input stroke belong to the same character according to the geometric position information of the current written stroke and the geometric position of the character composed of all the strokes that have been input before. The relationship before the information is judged.
  • the auxiliary judgment is performed in combination with the recognition result, and the specific method is: identifying the characters of all the strokes of the color A that have been written in addition to the current stroke, if the recognition confidence is recognized. If it is high, it is judged that the current stroke does not belong to the same character as the previous stroke;
  • the method of calculating the probability that the character to be recognized belongs to a candidate recognition result character is given by the recognition classifier;
  • the second decision function classifier is used as a classifier, and the probability is approximately equal to the negative exponential power of the recognition distance given by the second decision function classifier.
  • step S114 a method for determining whether the currently written stroke and all the strokes of the color A previously input belong to the same character include:
  • Step 141 Determine whether the current stroke is the first stroke input by the user; if yes, go to step 146; otherwise, go to step 142;
  • Step 142 determining whether the current stroke is a new written stroke on the right side of the previous stroke; if yes, go to step 145; otherwise, go to step 143;
  • Step 143 determining whether the current stroke overlaps with another stroke that has already been written; if yes, go to step 144; otherwise, go to step 145;
  • Step 144 determining that the current stroke overlaps with the already written stroke is greater than a given threshold; if yes, go to step 146; otherwise, go to step 145;
  • Step 145 returning a judgment result that the currently input stroke and the previously input stroke may belong to the same character
  • Step 146 returning the judgment result that the currently input stroke and the previously input stroke do not belong to the same character.
  • step S119 the method for determining that the recognition engine recognizes the recognition confidence of a handwritten character is: Calculating the difference between the probability that the character to be recognized belongs to the first candidate character and the consideration of the character to be recognized belonging to the second candidate character. If the difference is greater than a certain set threshold, the recognition result of the recognition result given by the recognition engine is considered high;
  • the method of calculating the probability that the character to be recognized belongs to a candidate recognition result character is given by the recognition classifier;
  • the second decision function classifier is used as a classifier, and the probability is approximately equal to the negative exponential power of the recognition distance given by the second decision function classifier.
  • a handwriting input method for overlapping writing includes the following steps: Step S111: Touching a touch screen through a touch unit to start inputting a stroke;
  • Step S112 The touch unit moves on the touch screen, records and displays the stroke of the stroke on the handwritten area on the touch screen, and draws the current stroke of the stroke with the given color A;
  • Step S113 The touch unit leaves the touch screen, and the input of the current stroke ends
  • Step S114 determining whether the currently written stroke and all the strokes of the color A that have been input before belong to the same character; if the answer is yes, keep the current stroke color unchanged; if the answer is no, go to step S115;
  • Step S115 determining whether there is a character of the inactive view layer on the screen; the character of the inactive view layer refers to the character that has been completely written and recognized and color-converted last time, and the character is given in step S117. If the answer is yes, go to step S116; if the answer is no, then go to step S117;
  • Step S116 Clearing a character image of the inactive view layer and related stroke information
  • Step S117 Defining a character composed of all strokes having a color value A other than the current stroke is an inactive view layer character; Go to step S118;
  • Step S118 Submit the current inactive view layer character data to the recognition engine for identification, and output the recognition result.
  • the method further includes after step S118: Step S119: determining whether the recognition result given by the recognition engine is sufficiently reliable; if the answer is yes, then proceeding to step S120; otherwise, proceeding to step S122; Step S120: gradually fade the stroke color of the currently inactive view layer character until it disappears and does not display;
  • Step S121 Clear the character image of the inactive view layer and the related stroke information; Go to step S123;
  • Step S122 Convert the stroke color of the current inactive view layer character to another color B for display; and then go to step S123;
  • Step S123 determining whether there is still a new stroke input, if the answer is yes, then proceeding to step S112; otherwise, proceeding to step S126;
  • Step S125 determine whether there is still a new stroke input, if the answer is yes, then proceeds to step S112, otherwise proceeds to step S126;
  • Step S126 submit the currently written character data to the recognition engine for identification, and output the recognition result
  • the device requires a general arithmetic and storage device, including a CPU of a certain frequency (central processing unit), a memory for calculation and a system software for storage, and a basic Operating system, application software and storage space for various data.
  • a touch screen capable of handwriting input is required in the device, and any area can be designated as a handwriting input area on the touch screen, and the touch screen also has a display area for identifying the result, a command button area, and the like.
  • the invention has the beneficial effects that: the handwriting input method of the overlapping writing proposed by the invention enables the user to perform overlapping display on the touch screen, and the proposed interface display processing method can make different overlapping characters display feedback to the user.
  • Figure 1 illustrates the effect of an interface display when writing the two characters "Scientific" in a conventional overlapping writing mode. (not processed by the method of the invention)
  • Figure 2 illustrates the effect of the screen interface display when writing the three characters "Scientist” in a conventional overlapping writing manner. (not processed by the method of the invention)
  • FIG. 3 is a flow chart of the handwriting input method of the present invention.
  • Figure 4 is a flow chart for determining whether the current stroke and the previous stroke belong to the same character step.
  • Figure 5 illustrates the effect of the screen interface display after processing the two characters "Scientific" in accordance with the method of the present invention.
  • Figure 6 illustrates the process of processing in accordance with the method of the present invention and the effect of the corresponding screen interface display when the three characters "Scientist" are written in an overlapping writing mode.
  • Figure 7 demonstrates the effect of the screen interface after the color gradient is processed when the first character " ⁇ " has a high recognition confidence when writing the two characters "Science” in an overlapping writing mode.
  • the device implementing the patent of the present invention can employ a smart phone with a touch screen (for example, HTC/Goog l e Nexus One smart phone) with a touch screen, which can collect the stroke track written by the user's finger.
  • a touch screen for example, HTC/Goog l e Nexus One smart phone
  • the invention can be well implemented by compiling various processing programs in the C++ language.
  • the present invention can also be implemented on other mobile electronic devices such as tablet computers, PDAs, GPS terminals, etc.
  • the present invention can also be implemented in other programming languages such as C language and Java language.
  • the present invention discloses a handwriting input method for overlapping writing, and the handwriting input method includes the following steps:
  • Step 1 00 Start, perform initialization processing: Clear the screen and clear the inactive view layer characters.
  • Step 1 1 0 The stylus or finger touches the touch screen and starts to input the stroke.
  • Step 120 The stylus or the finger moves on the touch screen, records and displays the stroke of the stroke on the handwriting area on the touch screen, and draws the current stroke of the stroke with the color value RGB (255, 0, 0).
  • Step 1 30 The stylus or finger leaves the touch screen, and the input of the current stroke ends.
  • Step 140 It is judged whether all the strokes of the currently written stroke and the previously input color are red (whose RGB values are taken as (255, 0, 0)) belong to the same character? If the answer is yes, go to the step 240. If the answer is no, go to step 150.
  • Judging whether the currently written stroke and the previously input stroke belong to the same character can be judged according to the relationship between the geometric position information of the current written stroke and the geometric position information of the characters composed of all the strokes that have been input before, a specific implementation
  • the flow chart is shown in Figure 4.
  • the specific method is: identifying the characters composed of all the strokes of the red (RGB ( 255, 0, 0 )) written in addition to the current stroke, If the recognition confidence is high, it is judged that the current stroke does not belong to the same character as the previous stroke.
  • the method for determining that the recognition engine recognizes the recognition confidence of a handwritten character is: calculating a difference between a probability that the character to be recognized belongs to the first candidate character and a consideration of the character to be recognized belonging to the second candidate character, if the difference is greater than a certain setting For a wide value (for example, 0.5), the recognition result given by the recognition engine is considered to be highly reliable.
  • the method for calculating the probability that the character to be recognized belongs to a candidate recognition result character is given by the recognition classifier.
  • the second decision function classifier is used as a classifier, and the probability is approximately equal to the second decision function classifier.
  • the negative exponential power of the recognition distance is: calculating a difference between a probability that the character to be recognized belongs to the first candidate character and a consideration of the character to be recognized belonging to the second candidate character, if the difference is greater than a certain setting For a wide value (for example, 0.5), the recognition result given by the recognition engine is considered to be highly reliable.
  • a method for judging whether the currently written stroke and all the strokes of the color A that have been input before belong to the same character include the following steps:
  • Step 141 Determine whether the current stroke is the first stroke input by the user; if yes, go to step 146; otherwise, go to step 142;
  • Step 142 determining whether the current stroke is a new written stroke on the right side of the previous stroke; if yes, go to step 145; otherwise, go to step 143;
  • Step 143 determining whether the current stroke overlaps with another stroke that has already been written; if yes, go to step 144; otherwise, go to step 145;
  • Step 144 determining that the current stroke overlaps with the already written stroke is greater than a given threshold; if yes, go to step 146; otherwise, go to step 145;
  • Step 145 returning a judgment result that the currently input stroke and the previously input stroke may belong to the same character
  • Step 146 returning the judgment result that the currently input stroke and the previously input stroke do not belong to the same character.
  • Step 150 Determine whether there are characters in the inactive view layer on the screen (so-called inactive view)
  • the character of the layer refers to the character that has been written completely and has been identified and color-converted, and the character is defined in step 170. If the answer is yes, go to step 160; if the answer is no, go to step 170. .
  • Step 160 Clear the character image of the inactive view layer and the related stroke information
  • Step 170 Defining a character consisting of all strokes except the current stroke with a red color value is an inactive view layer character
  • Step 180 submit the current inactive view layer character data to the recognition engine for identification, and output the recognition result
  • Step 190 Determine whether the recognition result given by the recognition engine is sufficiently reliable. If the answer is yes, then go to step 200; otherwise, go to step 220.
  • the method for determining that the recognition engine recognizes the recognition confidence of a handwritten character is: calculating a difference between a probability that the character to be recognized belongs to the first candidate character and a consideration of the character to be recognized belonging to the second candidate character, if the difference is greater than a certain setting For a wide value (for example, 0.5), the recognition result given by the recognition engine is considered to be highly reliable.
  • the method for calculating the probability that the character to be recognized belongs to a candidate recognition result character is given by the recognition classifier.
  • the second decision function classifier is used as a classifier, and the probability is approximately equal to the second decision function classifier.
  • the negative exponential power of the recognition distance is: calculating a difference between a probability that the character to be recognized belongs to the first candidate character and a consideration of the character to be recognized belonging to the second candidate character, if the difference is greater than a certain setting For a wide value (for example, 0.5), the recognition result given by the recognition engine is considered to be highly reliable.
  • Step 200 gradually fade the stroke color of the currently inactive view layer character until it disappears and does not display; in this embodiment, if the screen background color is white (RGB (255, 255, 255)), the specific implementation color
  • the pseudo-C language of the gradient is described as follows:
  • the stroke color is converted to RGB ( 255, i, j) ;
  • Step 210 Clear the character image of the inactive view layer and the related stroke information
  • Step 220 Convert the stroke color of the currently inactive view layer character to light gray (RGB (190, 190, 190) for display;
  • Figure 5 illustrates the effect of the screen interface displayed after the above steps are processed when the two characters "Science" are overlapped.
  • Step 230 judging whether there is still a new stroke input, if the answer is yes, waiting for the user to input a new stroke, go to step 120; otherwise, go to step 260;
  • Step 250 judging whether there is still a new stroke input, if the answer is yes, waiting for the user to input a new stroke, proceeds to step 120, otherwise proceeds to step 260;
  • Step 260 submit the currently written character data to the recognition engine for identification, and output the recognition result;
  • Fig. 6 illustrates the process of the above-described processing and the effect of the corresponding screen interface display in accordance with the method of the present invention when the three characters of "scientist" are written in an overlapping writing manner.
  • Embodiment 2
  • This embodiment discloses a handwriting input method for overlapping writing, and the handwriting input method includes the following steps:
  • Step S111 Touching the touch screen through the touch unit to start inputting a stroke
  • Step S112 The touch unit moves on the touch screen, records and displays the stroke track in the handwriting area on the touch screen, and draws the current stroke track with the given color A;
  • Step S113 The touch unit leaves the touch screen, and the input of the current stroke ends;
  • Step S114 determining whether the currently written stroke and all the strokes of the color A that have been input before belong to the same character; if the answer is yes, keep the current stroke color unchanged; if the answer is no, go to step S115;
  • Step S115 determining whether there is a character of the inactive view layer on the screen; the character of the inactive view layer refers to the character that has been completely written and recognized and color-converted last time, and the character is given in step S117. If the answer is yes, go to step S116; if the answer is no, then go to step S117;
  • Step S116 Clearing a character image of the inactive view layer and related stroke information
  • Step S117 Defining a character composed of all strokes having a color value A other than the current stroke is an inactive view layer character; Go to step S118;
  • Step S118 Submit the current inactive view layer character data to the recognition engine for identification, and output the recognition result.
  • the method further comprises after step S118:
  • Step S119 determining whether the recognition result given by the recognition engine is sufficiently reliable; if the answer is yes, then proceeding to step S120; otherwise, proceeding to step S122;
  • Step S120 gradually fade the stroke color of the currently inactive view layer character until it disappears and does not display
  • Step S121 Clear the character image of the inactive view layer and the related stroke information; Go to step S123;
  • Step S122 Convert the stroke color of the current inactive view layer character to another color B for display; and then go to step S123;
  • Step S123 determining whether there is still a new stroke input, if the answer is yes, then proceeding to step S112; otherwise, proceeding to step S126;
  • Step S125 determine whether there is still a new stroke input, if the answer is yes, then proceeds to step S112, otherwise proceeds to step S126;
  • Step S126 submit the currently written character data to the recognition engine for identification, and output the recognition result
  • the handwriting input method of the overlapping writing proposed by the present invention enables the user to display the interface display processing method on the touch screen, so that different overlapping characters can be well displayed and displayed to the user. It will not cause confusion of the display interface, thus ensuring that the user can perform continuous handwriting input according to the natural writing habit under this friendly interface, and can continuously write a plurality of handwritten characters without interruption, thereby effectively improving the efficiency of handwriting input.
  • the description and application of the present invention are intended to be illustrative, and not intended to limit the scope of the invention. Variations and modifications of the embodiments disclosed herein are possible, and various alternative and equivalent components of the embodiments are well known to those of ordinary skill in the art.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • General Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Human Computer Interaction (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Multimedia (AREA)
  • User Interface Of Digital Computer (AREA)
  • Character Discrimination (AREA)
  • Position Input By Displaying (AREA)

Description

一种重叠书写的手写输入方法 技术领域
本发明属于文字输入技术领域， 涉及一种手写输入方法， 尤其涉及一种重叠 书写的手写输入方法。
背景技术
近年来， 手写输入法作为一种重要的文字输入方式已经广泛应用于智能手 机、 平板电脑、 上网本、 GPS终端、 学习机等多种移动电子设备之中。 手写输 入法一般是指用户通过电子设备的触摸屏用书写笔或手指进行文字书写， 同时 电子设备对釆集到的文字书写轨迹进行识别， 转换为相应的文字并显示在屏幕 上。 在电子设备对用户书写轨迹的数据进行釆集的过程中， 一般都会实时将用 户的书写的轨迹显示在显示屏上， 以方便用户及时看到自己书写文字的笔画轨 迹反馈信息。
目前移动电子设备的触摸屏及显示屏的尺寸大小有限， 特别是在新一代以 iPhone及 Android手机为代表的智能手机， 其书写方式一般是感应人手指在电 容触摸屏的运动轨迹， 由于触摸屏的大小限制及手指触点的分辨率限制， 用户 一般一次只能在触摸屏上书写一到二个字符， 用户每书写完一个字符时， 通常 要停顿一段时间 （典型停留时间在 200ms ~ 1000ms左右） ， 然后等待移动设备 对上一次书写的字符进行检测及识别之后， 才能进行下一个字符的手写输入， 这样使得利用手写输入文字时效率低下， 无法按照人们自然书写的习惯进行连 续的手写输入。
为了提高书写输入效率， 有相关研究人员研发了全屏书写全屏识别的手写输 入方案， 一次可以在触摸屏上书写少量的多个字符， 然后再将所有书写的字符 图像提交给识别引擎进行分析及识别。 但该方法任然受限与触摸屏尺寸的大 小， 无法进行任意多字符的连续书写 （例如书写完一句完整的句子） ， 特别是 在触摸屏尺寸受限的移动电子设备（如智能手机）上， 通常触摸屏全屏能容纳 的书写字符数一般都很小 （例如 iPhone智能手机， 利用手指书写通常一个屏幕 能书写的字符数时小于 4个） ， 而且如果用户进行全屏书写时不同字符间出现 粘连或重叠情况， 将造成识别引擎无法进行正确分割和识别字符。
本发明利用重叠书写方式进行任意多字连续书写的手写文字输入方法， 可以 有效解决上述问题。 所谓重叠书写方式， 是指用户在写完一个字符后， 书写下 一个字符可以完全重叠或部分重叠在第一个字符之上的一种书写方式。 该方法 能让用户书写不同字符时不必停顿， 可连续在手写屏上书写多个手写字符并进 行相应的识别处理， 从而大大提高手写文字输入的效率。 但该方法的一个主要 问题是重叠书写时， 由于旧的字符笔画的存在， 会导致在触摸屏上显示新书写 字符笔画时受到旧字符笔画的干扰（如图 1所示 "科学" 两个字符的重叠书写 轨迹显示） ， 从而影响给用户的显示反馈， 给用户的新书写轨迹造成视觉干 扰， 影响用户体验及输入效率。 这一问题特别是在连续重叠书写多个字符时尤 其严重， 例如如图 3连续书写 "科学家" 这三个字符时， 在书写到第三个字符 "家" 时， 用户基本上看不清楚新书写的字符笔画轨迹。 发明内容
本发明所要解决的技术问题是： 提供一种重叠书写的手写输入方法， 通过判 断字符重叠的起始笔画， 对当前书写的重叠字符的笔画及上一次已经完的字符 笔画赋予不同的显示颜色， 从而达到显示美观清晰且不干扰用户进行手写输 入。
为解决上述技术问题， 本发明釆用如下技术方案：
一种重叠书写的手写输入方法， 所述手写输入方法包括如下步骤： 步骤 S111 : 手写笔或手指接触触摸屏， 开始输入笔画；
步骤 S112 : 手写笔或手指在触摸屏上移动， 记录并在触摸屏上的手写区域显 示笔画轨迹， 同时用给定的颜色 A绘制出当前笔画轨迹；
步骤 S11 3 : 手写笔或手指离开触摸屏， 当前笔画的输入结束；
步骤 S114 : 判断当前书写的笔画与之前已经输入的颜色为 A的所有笔画是否 属于同一个字符； 如果答案为是， 转到步骤 S124 ; 如果答案为否， 转到步骤
S115 ; 步骤 S115: 判断屏幕上是否存在有非激活视图层的字符； 所谓非激活视图层 的字符， 是指上一次已经书写完整且经过识别及颜色变换处理的字符， 该字符 在步骤 S117中给出定义； 如果答案为是， 转向步骤 S116; 如果答案为否， 则转 向步骤 S117;
步骤 S116: 清除非激活视图层的字符图像及与其相关笔画信息；
步骤 S117: 定义除当前笔画之外其余所有颜色值为 A的笔画所构成的字符为 非激活视图层字符；
步骤 S118: 将当前非激活视图层字符数据提交给识别引擎进行识别， 输出识 别结果；
步骤 S119: 判断识别引擎给出的识别结果是否足够可信； 如果答案为是， 则 转向步骤 S120; 否则转向步骤 S122;
步骤 S120: 将当前非激活视图层字符的笔画颜色逐渐变淡， 直到消失不显 示；
步骤 S121: 清除非激活视图层的字符图像及与其相关笔画信息； 转向步骤 S123;
步骤 S122: 将当前非激活视图层字符的笔画颜色变换为另外一种颜色 B进行 显示；
步骤 S123: 判断是否还有新的笔画输入， 如果答案为是， 则转向步骤 S112; 否则转向步骤 S126;
步骤 S124: 保持当前笔画颜色不变；
步骤 S125: ， 判断是否还有新的笔画输入， 如果答案为是， 则转入步骤 S112, 否则转入步骤 S126;
步骤 S126, 将当前所书写的字符数据提交给识别引擎进行识别， 并输出识别 结果；
步骤 S127, 结束。
作为本发明的一种优选方案， 步骤 S114中， 判断当前书写的笔画与之前已经 输入的笔画是否属于同一个字符根据当前书写笔画的几何位置信息与之前已经 输入的所有笔画组成的字符的几何位置信息之前的关系来进行判断。 作为本发明的一种优选方案， 步骤 S114中， 结合识别结果进行辅助判断， 具 体做法是： 对除了当前笔画之外已经书写的颜色为颜色 A所有笔画的构成的字 符进行识别， 如果识别置信度高， 则判断当前笔画与之前的笔画不属于同一个 字符；
判断识别引擎识别某个手写字符的识别置信度的办法是：
计算待识别字符属于首候选字符的概率与待识别字符属于第二候选字符的考 虑之差， 如果该差值大于某一个设定的阔值， 则认为该识别引擎给出的识别结 果可信度高；
计算待识别字符属于某个候选识别结果字符的概率的方法由识别分类器给 出；
以二次判决函数分类器作为分类器， 该概率近似等于二次判决函数分类器给 出的识别距离的负指数幂。
作为本发明的一种优选方案， 步骤 S114中， 判断当前书写的笔画与之前已经 输入的颜色为 A的所有笔画是否属于同一个字符的方法包括：
步骤 141 , 判断当前笔画是否为用户所输入的第一个笔画； 若是， 转至步骤 146 ; 否则转至步骤 142 ;
步骤 142 , 判断当前笔画是否是位于前一笔画右侧的新写文字笔画； 若是转 至步骤 145 ; 否则转至步骤 143 ;
步骤 143 , 判断当前笔画是否与别的已经书写的笔画重叠； 若是转至步骤 144 ; 否则转至步骤 145 ;
步骤 144 , 判断当前笔画与已经书写笔画的重叠程度大于某个给定阔值； 若 是转至步骤 146 ; 否则转至步骤 145 ;
步骤 145 , 返回当前输入的笔画与之前已经输入的笔画可能属于同一个字符 的判断结果；
步骤 146 , 返回当前输入的笔画与之前已经输入的笔画不属于同一个字符的 判断结果。
作为本发明的一种优选方案， 步骤 S119中， 判断识别引擎识别某个手写字符 的识别置信度的办法是： 计算待识别字符属于首候选字符的概率与待识别字符属于第二候选字符的考 虑之差， 如果该差值大于某一个设定的阔值， 则认为该识别引擎给出的识别结 果可信度高；
计算待识别字符属于某个候选识别结果字符的概率的方法由识别分类器给 出；
以二次判决函数分类器作为分类器， 该概率近似等于二次判决函数分类器给 出的识别距离的负指数幂。 一种重叠书写的手写输入方法， 所述手写输入方法包括如下步骤： 步骤 S111: 通过触摸单元接触触摸屏， 开始输入笔画；
步骤 S112: 触摸单元在触摸屏上移动， 记录并在触摸屏上的手写区域显示笔 画轨迹， 同时用给定的颜色 A绘制出当前笔画轨迹；
步骤 S113: 触摸单元离开触摸屏， 当前笔画的输入结束；
步骤 S114: 判断当前书写的笔画与之前已经输入的颜色为 A的所有笔画是否 属于同一个字符； 如果答案为是， 保持当前笔画颜色不变； 如果答案为否， 转 到步骤 S115;
步骤 S115: 判断屏幕上是否存在有非激活视图层的字符； 所谓非激活视图层 的字符， 是指上一次已经书写完整且经过识别及颜色变换处理的字符， 该字符 在步骤 S117中给出定义； 如果答案为是， 转向步骤 S116; 如果答案为否， 则转 向步骤 S117;
步骤 S116: 清除非激活视图层的字符图像及与其相关笔画信息；
步骤 S117: 定义除当前笔画之外其余所有颜色值为 A的笔画所构成的字符为 非激活视图层字符； 转至步骤 S118;
步骤 S118: 将当前非激活视图层字符数据提交给识别引擎进行识别， 输出识 别结果。
作为本发明的一种优选方案， 所述方法在步骤 S118之后进一步包括： 步骤 S119: 判断识别引擎给出的识别结果是否足够可信； 如果答案为是， 则 转向步骤 S120; 否则转向步骤 S122; 步骤 S120: 将当前非激活视图层字符的笔画颜色逐渐变淡， 直到消失不显 示；
步骤 S121: 清除非激活视图层的字符图像及与其相关笔画信息； 转向步骤 S123;
步骤 S122: 将当前非激活视图层字符的笔画颜色变换为另外一种颜色 B进行 显示； 转向步骤 S123;
步骤 S123: 判断是否还有新的笔画输入， 如果答案为是， 则转向步骤 S112; 否则转向步骤 S126;
步骤 S125: ， 判断是否还有新的笔画输入， 如果答案为是， 则转入步骤 S112, 否则转入步骤 S126;
步骤 S126, 将当前所书写的字符数据提交给识别引擎进行识别， 并输出识别 结果；
步骤 S127, 结束。
为了实行本发明， 需具备如下硬件条件： 该设备中需要有一般的运算和存储 装置， 包括一定频率的 CPU (中央处理器） ， 有一定用来运算的内存和用来存 储系统软件， 基本的操作系统， 应用软件和各种数据的存储空间等。 该设备中 需要有可进行手写输入的触摸屏， 该触摸屏上可以指定任意区域为手写输入区 域， 该触摸屏上还有识别结果的显示区域， 命令按钮区域等等。 本发明的有益效果在于： 本发明提出的重叠书写的手写输入方法， 使得用户 在触摸屏上进行重叠书写时， 所提出的界面显示处理方法可以使得不同的重叠 字符都能很好的反馈显示给用户， 而不会导致显示界面的混乱， 从而保证用户 可以在此友好的界面下按照自然书写的习惯进行连续的手写输入， 可以不间断 地连续书写多个手写文字， 从而有效的提高了手写输入的效率。 附图说明
图 1演示了一种常规重叠书写方式书写 "科学" 这两个字符时界面显示的效 果。 （未经过本发明方法进行处理） 图 2演示了按照常规重叠书写方式书写 "科学家" 这三个字符时屏幕界面显 示的效果。 （未经过本发明方法进行处理）
图 3是本发明手写输入方法的流程图。
图 4是判断当前书写笔画与之前笔画是否属于同一个字符步骤的流程图。 图 5演示了进行重叠书写 "科学" 这两个字符时， 按照本发明方法处理后屏 幕界面显示的效果。
图 6演示了重叠书写方式书写 "科学家" 这三个字符时， 按照本发明方法处 理的过程及相应的屏幕界面显示的效果。
图 7演示了重叠书写方式书写 "科学" 这两个字符时， 当第一个字符 "科" 的识别置信度高时， 其颜色渐变处理后的屏幕界面显示的效果。 具体实施方式
下面结合附图详细说明本发明的优选实施例。
实施例一
实施本发明专利的设备可以釆用带触摸屏的智能手机（例如 HTC/Goog l e Nexus One智能手机） ， 该手机带有触摸屏， 可以釆集用户手指书写的笔画轨 迹。 釆用 C++语言编制相应的各类处理程序， 便能很好的实施本发明。 本发明 也可以在平板电脑、 PDA、 GPS终端等其它移动电子设备上实现； 本发明也可以 釆用 C语言、 Java语言等其它编程语言进行实现。
请参阅图 3 , 本发明揭示了一种重叠书写的手写输入方法， 所述手写输入方 法包括如下步骤：
【步骤 1 00】开始， 进行初始化处理： 清空屏幕， 清空非激活视图层字符。
【步骤 1 1 0】手写笔或手指接触触摸屏， 开始输入笔画。
【步骤 120】手写笔或手指在触摸屏上移动， 记录并在触摸屏上的手写区域 显示笔画轨迹， 同时用颜色值 RGB ( 255, 0, 0 )绘制出当前笔画轨迹。
【步骤 1 30】手写笔或手指离开触摸屏， 当前笔画的输入结束。
【步骤 140】判断当前书写的笔画与之前已经输入的颜色为红色（其 RGB值取 为 （ 255, 0, 0 ) ) 的所有笔画是否属于同一个字符？ 如果答案为是， 转到步骤 240。 如果答案为否， 转到步骤 150。
判断当前书写的笔画与之前已经输入的笔画是否属于同一个字符可以根据当 前书写笔画的几何位置信息与之前已经输入的所有笔画组成的字符的几何位置 信息之前的关系来进行判断， 一个具体的实现流程图如图 4所示。 为了提高判 断准确性， 还可结合识别结果进行辅助判断， 具体做法是： 对除了当前笔画之 外已经书写的颜色为红色 （RGB ( 255, 0, 0 ) )所有笔画的构成的字符进行识 别， 如果识别置信度高， 则判断当前笔画与之前的笔画不属于同一个字符。
判断识别引擎识别某个手写字符的识别置信度的办法是： 计算待识别字符属 于首候选字符的概率与待识别字符属于第二候选字符的考虑之差， 如果该差值 大于某一个设定的阔值（例如 0. 5 ) , 则认为该识别引擎给出的识别结果可信 度高。 计算待识别字符属于某个候选识别结果字符的概率的方法由识别分类器 给出， 在本实施案例中， 以二次判决函数分类器作为分类器， 该概率近似等于 二次判决函数分类器给出的识别距离的负指数幂。
请参阅图 4 , 判断当前书写的笔画与之前已经输入的颜色为 A的所有笔画是否 属于同一个字符的方法包括如下步骤：
步骤 141 , 判断当前笔画是否为用户所输入的第一个笔画； 若是， 转至步骤 146 ; 否则转至步骤 142 ;
步骤 142 , 判断当前笔画是否是位于前一笔画右侧的新写文字笔画； 若是转 至步骤 145 ; 否则转至步骤 143 ;
步骤 143 , 判断当前笔画是否与别的已经书写的笔画重叠； 若是转至步骤 144 ; 否则转至步骤 145 ;
步骤 144 , 判断当前笔画与已经书写笔画的重叠程度大于某个给定阔值； 若 是转至步骤 146 ; 否则转至步骤 145 ;
步骤 145 , 返回当前输入的笔画与之前已经输入的笔画可能属于同一个字符 的判断结果；
步骤 146 , 返回当前输入的笔画与之前已经输入的笔画不属于同一个字符的 判断结果。
【步骤 150】判断屏幕上是否存在有非激活视图层的字符 （所谓非激活视图 层的字符， 是指上一次已经书写完整且经过识别及颜色变换处理的字符， 该 字符在步骤 170中给出定义） ， 如果答案为是， 转向步骤 160; 如果答案为否， 则转向步骤 170。
【步骤 160】清除非激活视图层的字符图像及与其相关笔画信息；
【步骤 170】定义除当前笔画之外其余所有颜色值为红色的笔画所构成的字 符为非激活视图层字符；
【步骤 180】将当前非激活视图层字符数据提交给识别引擎进行识别， 输出 识别结果；
【步骤 190】判断识别引擎给出的识别结果是否足够可信？ 如果答案为是， 则转向步骤 200; 否则转向步骤 220。
判断识别引擎识别某个手写字符的识别置信度的办法是： 计算待识别字符属 于首候选字符的概率与待识别字符属于第二候选字符的考虑之差， 如果该差值 大于某一个设定的阔值（例如 0. 5 ) , 则认为该识别引擎给出的识别结果可信 度高。 计算待识别字符属于某个候选识别结果字符的概率的方法由识别分类器 给出， 在本实施案例中， 以二次判决函数分类器作为分类器， 该概率近似等于 二次判决函数分类器给出的识别距离的负指数幂。
【步骤 200】将当前非激活视图层字符的笔画颜色逐渐变淡， 直到消失不显 示； 在本实施示例中， 设屏幕背景颜色为白色 （RGB ( 255, 255, 255 ) ) , 则具 体实现颜色渐变的伪 C语言描述如下：
{
定义颜色变量 i=0;
定义颜色变量 j=0;
whi le (i<= 255 && j<=255)
{
笔画颜色变换为 RGB ( 255, i, j) ;
i = i+10;
j = j+10;
显示笔画颜色； 延时 lms ;
具体的实现效果如图 7所示。
【步骤 210】清除非激活视图层的字符图像及与其相关笔画信息； 转向步骤
230;
【步骤 220】将当前非激活视图层字符的笔画颜色变换为浅灰色 （RGB ( 190, 190, 190 )进行显示；
图 5演示了进行重叠书写 "科学" 这两个字符时， 按照上述步骤处理后屏幕 界面显示的效果。
【步骤 230】判断是否还有新的笔画输入， 如果答案为是， 则等待用户输入 新的笔画， 转向步骤 120; 否则转向步骤 260;
【步骤 240】保持当前笔画颜色不变；
【步骤 250】判断是否还有新的笔画输入， 如果答案为是， 则等待用户输入 新的笔画， 转入步骤 120 , 否则转入步骤 260;
【步骤 260】将当前所书写的字符数据提交给识别引擎进行识别， 并输出识 别结果；
【步骤 270】结束。
图 6演示了重叠书写方式书写 "科学家" 这三个字符时， 按照本发明方法所 述的上述处理的过程及相应的屏幕界面显示的效果。 实施例二
本实施例揭示一种重叠书写的手写输入方法， 所述手写输入方法包括如下步 骤：
步骤 S111 : 通过触摸单元接触触摸屏， 开始输入笔画；
步骤 S112 : 触摸单元在触摸屏上移动， 记录并在触摸屏上的手写区域显示笔 画轨迹， 同时用给定的颜色 A绘制出当前笔画轨迹；
步骤 S113: 触摸单元离开触摸屏， 当前笔画的输入结束； 步骤 S114: 判断当前书写的笔画与之前已经输入的颜色为 A的所有笔画是否 属于同一个字符； 如果答案为是， 保持当前笔画颜色不变； 如果答案为否， 转 到步骤 S115;
步骤 S115: 判断屏幕上是否存在有非激活视图层的字符； 所谓非激活视图层 的字符， 是指上一次已经书写完整且经过识别及颜色变换处理的字符， 该字符 在步骤 S117中给出定义； 如果答案为是， 转向步骤 S116; 如果答案为否， 则转 向步骤 S117;
步骤 S116: 清除非激活视图层的字符图像及与其相关笔画信息；
步骤 S117: 定义除当前笔画之外其余所有颜色值为 A的笔画所构成的字符为 非激活视图层字符； 转至步骤 S118;
步骤 S118: 将当前非激活视图层字符数据提交给识别引擎进行识别， 输出识 别结果。
优选地， 所述方法在步骤 S118之后进一步包括：
步骤 S119: 判断识别引擎给出的识别结果是否足够可信； 如果答案为是， 则 转向步骤 S120; 否则转向步骤 S122;
步骤 S120: 将当前非激活视图层字符的笔画颜色逐渐变淡， 直到消失不显 示；
步骤 S121: 清除非激活视图层的字符图像及与其相关笔画信息； 转向步骤 S123;
步骤 S122: 将当前非激活视图层字符的笔画颜色变换为另外一种颜色 B进行 显示； 转向步骤 S123;
步骤 S123: 判断是否还有新的笔画输入， 如果答案为是， 则转向步骤 S112; 否则转向步骤 S126;
步骤 S125: ， 判断是否还有新的笔画输入， 如果答案为是， 则转入步骤 S112, 否则转入步骤 S126;
步骤 S126, 将当前所书写的字符数据提交给识别引擎进行识别， 并输出识别 结果；
步骤 S127, 结束。 综上所述， 本发明提出的重叠书写的手写输入方法， 使得用户在触摸屏上进 行重叠书写时， 所提出的界面显示处理方法可以使得不同的重叠字符都能很好 的反馈显示给用户， 而不会导致显示界面的混乱， 从而保证用户可以在此友好 的界面下按照自然书写的习惯进行连续的手写输入， 可以不间断地连续书写多 个手写文字， 从而有效的提高了手写输入的效率。 这里本发明的描述和应用是说明性的， 并非想将本发明的范围限制在上述实 施例中。 这里所披露的实施例的变形和改变是可能的， 对于那些本领域的普通 技术人员来说实施例的替换和等效的各种部件是公知的。 本领域技术人员应该 清楚的是， 在不脱离本发明的精神或本质特征的情况下， 本发明可以以其它形 式、 结构、 布置、 比例， 以及用其它组件、 材料和部件来实现。 在不脱离本发 明范围和精神的情况下， 可以对这里所披露的实施例进行其它变形和改变。

Claims

权利要求书
1、 一种重叠书写的手写输入方法， 其特征在于， 所述手写输入方法包括如下 步骤：
步骤 100: 进行初始化处理： 清空屏幕， 清空非激活视图层字符；
步骤 110: 手写笔或手指接触触摸屏， 开始输入笔画；
步骤 120: 手写笔或手指在触摸屏上移动， 记录并在触摸屏上的手写区域显 示笔画轨迹， 同时用第一颜色值绘制出当前笔画轨迹；
步骤 1 30: 手写笔或手指离开触摸屏， 当前笔画的输入结束；
步骤 140: 判断当前书写的笔画与之前已经输入的颜色为第一颜色值的所有 笔画是否属于同一个字符； 如果答案为是， 转到步骤 240; 如果答案为否， 转 到步骤 150;
判断当前书写的笔画与之前已经输入的笔画是否属于同一个字符根据当前书 写笔画的几何位置信息与之前已经输入的所有笔画组成的字符的几何位置信息 之前的关系来进行判断；
为了提高判断准确性， 结合识别结果进行辅助判断， 具体做法是： 对除了当 前笔画之外已经书写的颜色为第一颜色值所有笔画的构成的字符进行识别， 如 果识别置信度高， 则判断当前笔画与之前的笔画不属于同一个字符；
判断识别引擎识别某个手写字符的识别置信度的办法是： 计算待识别字符属 于首候选字符的概率与待识别字符属于第二候选字符的考虑之差， 如果该差值 大于某一个设定的阔值， 则认为该识别引擎给出的识别结果可信度高； 计算待 识别字符属于某个候选识别结果字符的概率的方法由识别分类器给出； 以二次 判决函数分类器作为分类器， 该概率近似等于二次判决函数分类器给出的识别 距离的负指数幂；
步骤 140包括如下步骤：
步骤 141 , 判断当前笔画是否为用户所输入的第一个笔画； 若是， 转至步骤 146 ; 否则转至步骤 142 ;
步骤 142 , 判断当前笔画是否是位于前一笔画右侧的新写文字笔画； 若是转 至步骤 145 ; 否则转至步骤 143 ;
步骤 143 , 判断当前笔画是否与别的已经书写的笔画重叠； 若是转至步骤 144 ; 否则转至步骤 145 ;
步骤 144 , 判断当前笔画与已经书写笔画的重叠程度大于某个给定阔值； 若 是转至步骤 146 ; 否则转至步骤 145 ;
步骤 145 , 返回当前输入的笔画与之前已经输入的笔画可能属于同一个字符 的判断结果；
步骤 146 , 返回当前输入的笔画与之前已经输入的笔画不属于同一个字符的 判断结果；
步骤 150: 判断屏幕上是否存在有非激活视图层的字符； 所述非激活视图层 的字符， 是指上一次已经书写完整且经过识别及颜色变换处理的字符， 该字符 在步骤 170中给出定义； 如果答案为是， 转向步骤 160; 如果答案为否， 则转向 步骤 170;
步骤 160： 清除非激活视图层的字符图像及与其相关笔画信息；
步骤 170: 定义除当前笔画之外其余所有颜色值为第一颜色值的笔画所构成 的字符为非激活视图层字符；
步骤 180: 将当前非激活视图层字符数据提交给识别引擎进行识别， 输出识 别结果；
步骤 190: 判断识别引擎给出的识别结果是否足够可信； 如果答案为是， 则 转向步骤 200; 否则转向步骤 220;
判断识别引擎识别某个手写字符的识别置信度的办法是： 计算待识别字符属 于首候选字符的概率与待识别字符属于第二候选字符的考虑之差， 如果该差值 大于某一个设定的阔值， 则认为该识别引擎给出的识别结果可信度高； 计算待 识别字符属于某个候选识别结果字符的概率的方法由识别分类器给出； 以二次 判决函数分类器作为分类器， 该概率近似等于二次判决函数分类器给出的识别 距离的负指数幂；
步骤 200: 将当前非激活视图层字符的笔画颜色逐渐变淡， 直到消失不显 步骤 210: 清除非激活视图层的字符图像及与其相关笔画信息； 转向步骤 230;
步骤 220: 将当前非激活视图层字符的笔画颜色变换为第二颜色进行显示； 步骤 230: 判断是否还有新的笔画输入， 如果答案为是， 则等待用户输入新 的笔画， 转向步骤 120; 否则转向步骤 260;
步骤 240: 保持当前笔画颜色不变；
步骤 250: 判断是否还有新的笔画输入， 如果答案为是， 则等待用户输入新 的笔画， 转入步骤 120 , 否则转入步骤 260;
步骤 260 , 将当前所书写的字符数据提交给识别引擎进行识别， 并输出识别 结果；
步骤 270 , 结束。
2、 一种重叠书写的手写输入方法， 其特征在于， 所述手写输入方法包括如下 步骤：
步骤 S111 : 手写笔或手指接触触摸屏， 开始输入笔画；
步骤 S112 : 手写笔或手指在触摸屏上移动， 记录并在触摸屏上的手写区域显 示笔画轨迹， 同时用给定的颜色 A绘制出当前笔画轨迹；
步骤 S11 3 : 手写笔或手指离开触摸屏， 当前笔画的输入结束；
步骤 S114 : 判断当前书写的笔画与之前已经输入的颜色为 A的所有笔画是否 属于同一个字符； 如果答案为是， 转到步骤 S124 ; 如果答案为否， 转到步骤
S115 ;
步骤 S115 : 判断屏幕上是否存在有非激活视图层的字符； 所谓非激活视图层 的字符， 是指上一次已经书写完整且经过识别及颜色变换处理的字符， 该字符 在步骤 S117中给出定义； 如果答案为是， 转向步骤 S116 ; 如果答案为否， 则转 向步骤 S117 ;
步骤 S116 : 清除非激活视图层的字符图像及与其相关笔画信息；
步骤 S117 : 定义除当前笔画之外其余所有颜色值为 A的笔画所构成的字符为 非激活视图层字符； 步骤 S118: 将当前非激活视图层字符数据提交给识别引擎进行识别， 输出识 别结果；
步骤 S119: 判断识别引擎给出的识别结果是否足够可信； 如果答案为是， 则 转向步骤 S120; 否则转向步骤 S122;
步骤 S120: 将当前非激活视图层字符的笔画颜色逐渐变淡， 直到消失不显 示；
步骤 S121: 清除非激活视图层的字符图像及与其相关笔画信息； 转向步骤 S123;
步骤 S122: 将当前非激活视图层字符的笔画颜色变换为另外一种颜色 B进行 显示；
步骤 S123: 判断是否还有新的笔画输入， 如果答案为是， 则转向步骤 S112; 否则转向步骤 S126;
步骤 S124: 保持当前笔画颜色不变；
步骤 S125: ， 判断是否还有新的笔画输入， 如果答案为是， 则转入步骤 S112, 否则转入步骤 S126;
步骤 S126, 将当前所书写的字符数据提交给识别引擎进行识别， 并输出识别 结果；
步骤 S127, 结束。
3、 根据权利要求 2所述的重叠书写的手写输入方法， 其特征在于：
步骤 S114中， 判断当前书写的笔画与之前已经输入的笔画是否属于同一个字 符根据当前书写笔画的几何位置信息与之前已经输入的所有笔画组成的字符的 几何位置信息之前的关系来进行判断。
4、 根据权利要求 2所述的重叠书写的手写输入方法， 其特征在于：
步骤 S114中， 结合识别结果进行辅助判断， 具体做法是： 对除了当前笔画之 外已经书写的颜色为颜色 A所有笔画的构成的字符进行识别， 如果识别置信度 高， 则判断当前笔画与之前的笔画不属于同一个字符； 判断识别引擎识别某个手写字符的识别置信度的办法是：
计算待识别字符属于首候选字符的概率与待识别字符属于第二候选字符的考 虑之差， 如果该差值大于某一个设定的阔值， 则认为该识别引擎给出的识别结 果可信度高；
计算待识别字符属于某个候选识别结果字符的概率的方法由识别分类器给 出；
以二次判决函数分类器作为分类器， 该概率近似等于二次判决函数分类器给 出的识别距离的负指数幂。
5、 根据权利要求 2所述的重叠书写的手写输入方法， 其特征在于：
步骤 S114中， 判断当前书写的笔画与之前已经输入的颜色为 A的所有笔画是 否属于同一个字符的方法包括：
步骤 141 , 判断当前笔画是否为用户所输入的第一个笔画； 若是， 转至步骤 146 ; 否则转至步骤 142 ;
步骤 142 , 判断当前笔画是否是位于前一笔画右侧的新写文字笔画； 若是转 至步骤 145 ; 否则转至步骤 143 ;
步骤 143 , 判断当前笔画是否与别的已经书写的笔画重叠； 若是转至步骤 144 ; 否则转至步骤 145 ;
步骤 144 , 判断当前笔画与已经书写笔画的重叠程度大于某个给定阔值； 若 是转至步骤 146 ; 否则转至步骤 145 ;
步骤 145 , 返回当前输入的笔画与之前已经输入的笔画可能属于同一个字符 的判断结果；
步骤 146 , 返回当前输入的笔画与之前已经输入的笔画不属于同一个字符的 判断结果。
6、 根据权利要求 2所述的重叠书写的手写输入方法， 其特征在于：
步骤 SI 19中， 判断识别引擎识别某个手写字符的识别置信度的办法是： 计算待识别字符属于首候选字符的概率与待识别字符属于第二候选字符的考 虑之差， 如果该差值大于某一个设定的阔值， 则认为该识别引擎给出的识别结 果可信度高；
计算待识别字符属于某个候选识别结果字符的概率的方法由识别分类器给 出；
以二次判决函数分类器作为分类器， 该概率近似等于二次判决函数分类器给 出的识别距离的负指数幂。
7、 一种重叠书写的手写输入方法， 其特征在于， 所述手写输入方法包括： 步骤 S111 : 通过触摸单元接触触摸屏， 开始输入笔画；
步骤 S112 : 触摸单元在触摸屏上移动， 记录并在触摸屏上的手写区域显示笔 画轨迹， 同时用给定的颜色 A绘制出当前笔画轨迹；
步骤 S11 3 : 触摸单元离开触摸屏， 当前笔画的输入结束；
步骤 S114 : 判断当前书写的笔画与之前已经输入的颜色为 A的所有笔画是否 属于同一个字符； 如果答案为是， 保持当前笔画颜色不变； 如果答案为否， 转 到步骤 S115 ;
步骤 S115 : 判断屏幕上是否存在有非激活视图层的字符； 所谓非激活视图层 的字符， 是指上一次已经书写完整且经过识别及颜色变换处理的字符， 该字符 在步骤 S117中给出定义； 如果答案为是， 转向步骤 S116 ; 如果答案为否， 则转 向步骤 S117 ;
步骤 S116 : 清除非激活视图层的字符图像及与其相关笔画信息；
步骤 S117 : 定义除当前笔画之外其余所有颜色值为 A的笔画所构成的字符为 非激活视图层字符； 转至步骤 S 118 ;
步骤 S118 : 将当前非激活视图层字符数据提交给识别引擎进行识别， 输出识 别结果。
8、 根据权利要求 7所述的重叠书写的手写输入方法， 其特征在于：
所述方法在步骤 S 118之后进一步包括：
步骤 S119 : 判断识别引擎给出的识别结果是否足够可信； 如果答案为是， 则 转向步骤 S 120; 否则转向步骤 S 122 ;
步骤 S120: 将当前非激活视图层字符的笔画颜色逐渐变淡， 直到消失不显 示；
步骤 S121 : 清除非激活视图层的字符图像及与其相关笔画信息； 转向步骤 S123;
步骤 S122 : 将当前非激活视图层字符的笔画颜色变换为另外一种颜色 B进行 显示， 或者将新输入的笔画的颜色变换为另外一种颜色 B; 转向步骤 S123; 步骤 S123 : 判断是否还有新的笔画输入， 如果答案为是， 则转向步骤 S 112 ; 否则转向步骤 S126 ;
步骤 S125 : ， 判断是否还有新的笔画输入， 如果答案为是， 则转入步骤 S112 , 否则转入步骤 S126 ;
步骤 S126 , 将当前所书写的字符数据提交给识别引擎进行识别， 并输出识别 结果；
步骤 S127 , 结束。
9、 根据权利要求 8所述的重叠书写的手写输入方法， 其特征在于：
步骤 SI 19中， 判断识别引擎识别某个手写字符的识别置信度的办法是： 计算待识别字符属于首候选字符的概率与待识别字符属于第二候选字符的考 虑之差， 如果该差值大于某一个设定的阔值， 则认为该识别引擎给出的识别结 果可信度高；
计算待识别字符属于某个候选识别结果字符的概率的方法由识别分类器给 出；
以二次判决函数分类器作为分类器， 该概率近似等于二次判决函数分类器给 出的识别距离的负指数幂。
10、 根据权利要求 7所述的重叠书写的手写输入方法， 其特征在于：
步骤 S114中， 判断当前书写的笔画与之前已经输入的笔画是否属于同一个字 符根据当前书写笔画的几何位置信息与之前已经输入的所有笔画组成的字符的 几何位置信息之前的关系来进行判断；
同时结合识别结果进行辅助判断， 具体做法是： 对除了当前笔画之外已经书 写的颜色为颜色 A所有笔画的构成的字符进行识别， 如果识别置信度高， 则判 断当前笔画与之前的笔画不属于同一个字符；
判断识别引擎识别某个手写字符的识别置信度的办法是： 计算待识别字符属 于首候选字符的概率与待识别字符属于第二候选字符的考虑之差， 如果该差值 大于某一个设定的阔值， 则认为该识别引擎给出的识别结果可信度高； 计算待 识别字符属于某个候选识别结果字符的概率的方法由识别分类器给出； 以二次 判决函数分类器作为分类器， 该概率近似等于二次判决函数分类器给出的识别 距离的负指数幂；
步骤 S114中， 判断当前书写的笔画与之前已经输入的颜色为 A的所有笔画是 否属于同一个字符的方法包括：
步骤 141 , 判断当前笔画是否为用户所输入的第一个笔画； 若是， 转至步骤 146 ; 否则转至步骤 142 ;
步骤 142 , 判断当前笔画是否是位于前一笔画右侧的新写文字笔画； 若是转 至步骤 145 ; 否则转至步骤 143 ;
步骤 143 , 判断当前笔画是否与别的已经书写的笔画重叠； 若是转至步骤 144 ; 否则转至步骤 145 ;
步骤 144 , 判断当前笔画与已经书写笔画的重叠程度大于某个给定阔值； 若 是转至步骤 146 ; 否则转至步骤 145 ;
步骤 145 , 返回当前输入的笔画与之前已经输入的笔画可能属于同一个字符 的判断结果；
步骤 146 , 返回当前输入的笔画与之前已经输入的笔画不属于同一个字符的 判断结果。
PCT/CN2011/077950 2010-12-10 2011-08-03 一种重叠书写的手写输入方法 WO2012075821A1 (zh)

Priority Applications (4)

Application Number Priority Date Filing Date Title
EP11846641.6A EP2650767A4 (en) 2010-12-10 2011-08-03 OVERLAPPED HANDWRITING ENTRY PROCEDURE
KR1020137014847A KR101825154B1 (ko) 2010-12-10 2011-08-03 중첩 쓰기 핸드라이팅 입력방법
US13/992,392 US9052755B2 (en) 2010-12-10 2011-08-03 Overlapped handwriting input method
JP2013542348A JP5784141B2 (ja) 2010-12-10 2011-08-03 重畳筆記による手書き入力方法

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN2010105836011A CN102156609B (zh) 2010-12-10 2010-12-10 一种重叠书写的手写输入方法
CN201010583601.1 2010-12-10

Publications (1)

Publication Number Publication Date
WO2012075821A1 true WO2012075821A1 (zh) 2012-06-14

Family

ID=44438124

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2011/077950 WO2012075821A1 (zh) 2010-12-10 2011-08-03 一种重叠书写的手写输入方法

Country Status (6)

Country Link
US (1) US9052755B2 (zh)
EP (1) EP2650767A4 (zh)
JP (1) JP5784141B2 (zh)
KR (1) KR101825154B1 (zh)
CN (1) CN102156609B (zh)
WO (1) WO2012075821A1 (zh)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103513898A (zh) * 2012-06-21 2014-01-15 夏普株式会社 手写字符切分方法和电子设备
WO2014099948A1 (en) * 2012-12-17 2014-06-26 Microsoft Corporation Composition of handwritten messages on mobile computing devices
US9881224B2 (en) 2013-12-17 2018-01-30 Microsoft Technology Licensing, Llc User interface for overlapping handwritten text input

Families Citing this family (22)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102508598B (zh) * 2011-10-09 2014-03-05 北京捷通华声语音技术有限公司 一种字符笔画渐变消隐方法及装置
CN102855017B (zh) * 2012-07-25 2015-07-08 东莞宇龙通信科技有限公司 终端和防止终端误输入的方法
US10254952B2 (en) * 2012-09-26 2019-04-09 Google Llc Progress display of handwriting input
CN103049200B (zh) * 2012-12-17 2018-08-17 上海海知信息技术有限公司 一种双画布映射实现方法
CN103049211B (zh) * 2012-12-17 2018-08-17 上海海知信息技术有限公司 一种手写输入法的界面分布方法及输入法
CN103064624A (zh) * 2012-12-27 2013-04-24 深圳市汇顶科技股份有限公司 触摸终端及其屏幕激活方法和系统
CN103309520A (zh) * 2013-05-20 2013-09-18 南京恒知讯科技有限公司 一种屏幕操作轨迹和声音输入同步存储处理方法、系统及终端
CN103345364B (zh) * 2013-07-09 2016-01-27 广东欧珀移动通信有限公司 电子手绘方法与系统
CN103345365B (zh) * 2013-07-12 2016-04-13 北京蒙恬科技有限公司 连续手写输入的显示方法及采用该方法的手写输入装置
JP2015069284A (ja) * 2013-09-27 2015-04-13 株式会社リコー 画像処理装置
JP6447066B2 (ja) * 2014-12-01 2019-01-09 株式会社リコー 画像処理装置、画像処理方法、及びプログラム
JP2016134014A (ja) * 2015-01-20 2016-07-25 株式会社リコー 電子情報ボード装置、情報処理方法およびプログラム
US10627244B1 (en) 2015-04-29 2020-04-21 Amazon Technologies, Inc. Image assisted delivery
US9904847B2 (en) * 2015-07-10 2018-02-27 Myscript System for recognizing multiple object input and method and product for same
CN105825535B (zh) * 2016-03-18 2018-06-05 广东欧珀移动通信有限公司 一种图层合并方法及系统
KR102551568B1 (ko) 2016-07-11 2023-07-06 삼성전자주식회사 전자 장치 및 그 제어 방법
CN109697905A (zh) * 2017-10-20 2019-04-30 深圳市鹰硕技术有限公司 一种试卷批阅系统
CN109992123A (zh) * 2018-01-02 2019-07-09 北京搜狗科技发展有限公司 输入方法、装置和机器可读介质
KR20200078932A (ko) * 2018-12-24 2020-07-02 삼성전자주식회사 전자 장치 및 전자 장치의 제어 방법
CN112396032B (zh) * 2020-12-03 2023-04-07 北京有竹居网络技术有限公司 书写检测方法、装置、存储介质及电子设备
CN112699780A (zh) * 2020-12-29 2021-04-23 上海臣星软件技术有限公司 对象识别方法、装置、设备及存储介质
CN113326727A (zh) * 2021-03-05 2021-08-31 临沂大学 一种书法教学用演示方法和装置

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1272691C (zh) * 2003-09-29 2006-08-30 摩托罗拉公司 用于手写的连写字符的识别方法及其装置
CN101620501A (zh) * 2009-07-28 2010-01-06 上海合合信息科技发展有限公司 手写输入系统及方法、电子设备

Family Cites Families (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPS60205686A (ja) * 1984-03-30 1985-10-17 Hitachi Ltd 手書き文字・図形認識システム
JPH08202826A (ja) * 1995-01-25 1996-08-09 Matsushita Electric Ind Co Ltd 携帯端末装置
JPH08286808A (ja) 1995-04-18 1996-11-01 Canon Inc 軌跡入出力電子装置及びその表示制御方法
CN1173247C (zh) * 1999-01-13 2004-10-27 国际商业机器公司 具有文字分割用户接口的手写信息处理系统
CN1139990C (zh) 1999-04-29 2004-02-25 华邦电子股份有限公司 使用者专用的非标准集成电路及其制造方法
JP3970075B2 (ja) * 2002-03-29 2007-09-05 三洋電機株式会社 文字認識装置、文字認識方法、その実行プログラムおよびそれを記録した記録媒体
JP4631005B2 (ja) * 2005-07-15 2011-02-16 公立大学法人大阪府立大学 付加情報を用いたパターン認識に係るプログラム、装置および方法
US20080008387A1 (en) * 2006-07-06 2008-01-10 Cheng Yi-Hsun E Method and apparatus for recognition of handwritten symbols
JP2009289188A (ja) * 2008-05-30 2009-12-10 Nec Corp 文字入力装置、文字入力方法および文字入力用プログラム
CN101539822B (zh) * 2009-01-21 2011-05-25 宇龙计算机通信科技(深圳)有限公司 一种触摸屏手写区域识别方法、系统
US8310461B2 (en) * 2010-05-13 2012-11-13 Nuance Communications Inc. Method and apparatus for on-top writing
CN101893987A (zh) * 2010-06-01 2010-11-24 华南理工大学 一种电子设备的手写输入方法

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1272691C (zh) * 2003-09-29 2006-08-30 摩托罗拉公司 用于手写的连写字符的识别方法及其装置
CN101620501A (zh) * 2009-07-28 2010-01-06 上海合合信息科技发展有限公司 手写输入系统及方法、电子设备

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP2650767A4 *

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103513898A (zh) * 2012-06-21 2014-01-15 夏普株式会社 手写字符切分方法和电子设备
WO2014099948A1 (en) * 2012-12-17 2014-06-26 Microsoft Corporation Composition of handwritten messages on mobile computing devices
US9846536B2 (en) 2012-12-17 2017-12-19 Microsoft Technology Licensing, Llc Composition of handwritten messages on mobile computing devices
US9881224B2 (en) 2013-12-17 2018-01-30 Microsoft Technology Licensing, Llc User interface for overlapping handwritten text input

Also Published As

Publication number Publication date
US20150030249A1 (en) 2015-01-29
JP5784141B2 (ja) 2015-09-24
KR101825154B1 (ko) 2018-02-02
CN102156609A (zh) 2011-08-17
EP2650767A1 (en) 2013-10-16
CN102156609B (zh) 2012-12-19
JP2014502399A (ja) 2014-01-30
EP2650767A4 (en) 2013-11-27
US9052755B2 (en) 2015-06-09
KR20130137648A (ko) 2013-12-17

Similar Documents

Publication Publication Date Title
WO2012075821A1 (zh) 一种重叠书写的手写输入方法
WO2012075820A1 (zh) 多字符连续书写的手写输入方法
US8355007B2 (en) Methods for use with multi-touch displays for determining when a touch is processed as a mouse event
US20160011682A1 (en) Touchpad operational mode
WO2017143860A1 (zh) 基于交互式电子白板的触摸操作方法及其系统
US9569107B2 (en) Gesture keyboard with gesture cancellation
US10082888B2 (en) Stylus modes
US9612697B2 (en) Touch control method of capacitive and electromagnetic dual-mode touch screen and handheld electronic device
JP2010102395A (ja) 画像処理装置、画像処理方法およびプログラム
WO2022257870A1 (zh) 一种虚拟标尺显示方法以及相关设备
CN110850982A (zh) 基于ar的人机交互学习方法、系统、设备及存储介质
EP4379521A1 (en) Writing/drawing content display method and related device
US10521108B2 (en) Electronic apparatus for detecting touch, method of controlling the same, and display apparatus including touch controller
CN103809794A (zh) 一种信息处理方法以及电子设备
TW201734709A (zh) 可切換輸入法的電子裝置及其輸入法切換方法、系統
US11561649B2 (en) Operation judgment method for interactive touch system
KR20240062249A (ko) 전자 장치 및 그 터치 좌표 획득 방법
TWI557608B (zh) 電子裝置的輸入裝置及其設定方法
CN105528059A (zh) 一种三维空间手势操作方法及系统
WO2019084876A1 (zh) 针对写字评分系统的判定方法、系统及装置
KR20110128220A (ko) 터치패드를 구비한 단말에서의 연산방법

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 11846641

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 2013542348

Country of ref document: JP

Kind code of ref document: A

WWE Wipo information: entry into national phase

Ref document number: 13992392

Country of ref document: US

ENP Entry into the national phase

Ref document number: 20137014847

Country of ref document: KR

Kind code of ref document: A

NENP Non-entry into the national phase

Ref country code: DE

WWE Wipo information: entry into national phase

Ref document number: 2011846641

Country of ref document: EP