CN110070080A - A kind of character detecting method and device, equipment and computer readable storage medium - Google Patents
A kind of character detecting method and device, equipment and computer readable storage medium Download PDFInfo
- Publication number
- CN110070080A CN110070080A CN201910184219.4A CN201910184219A CN110070080A CN 110070080 A CN110070080 A CN 110070080A CN 201910184219 A CN201910184219 A CN 201910184219A CN 110070080 A CN110070080 A CN 110070080A
- Authority
- CN
- China
- Prior art keywords
- image
- pixels
- block
- detected
- default
- Prior art date
- Legal status (The legal status 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 status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/255—Detecting or recognising potential candidate objects based on visual cues, e.g. shapes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/107—Static hand or arm
-
- 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
Abstract
This application discloses a kind of character detecting method and devices, this method comprises: obtaining image to be detected of object to be detected, and the image to be detected is converted into the image containing hue information, wherein the pixel in the image with hue information is respectively provided with corresponding tone value；The tone value of pixel is compared with default tone value range, obtains at least one block of pixels being made of pixel of the tone value within the scope of default tone value；Selection meets the block of pixels of preset condition from least one described block of pixels, determines text detection region according to the position of selected block of pixels.Using this method, text detection region can be quickly and accurately identified, removal does not need the region of text detection, improves text detection speed, improves text detection accuracy, improves user experience.Also disclose corresponding equipment and computer readable storage medium.
Description
Technical field
This application involves text detection technical field more particularly to a kind of character detecting methods and device, equipment and calculating
Machine readable storage medium storing program for executing.
Background technique
Currently, there is text detection demand in many application fields, and it is desirable that text detection is not only quickly but also accurate, and
And the power consumption of text detection equipment is again low.Text detection technology in the prior art is specifically included that based on traditional computer vision
Text detection, and the deep learning text detection based on artificial intelligence.But in practical applications, traditional calculations are based on
The text detection mode of machine vision is influenced by factors such as light, noise, background pictures, the accuracy rate of text detection and identification
It is very low；And the deep learning text detection mode based on artificial intelligence uses and is based on convolutional neural networks (CNN), although most of
In the case of text detection and the accuracy rate of identification it is higher, but its operation is high to hardware resource requirements, the speed of service is relatively slow, power consumption
Height, and text detection equipment needs the support of image processing unit (GPU), however mobile phone, intelligent glasses etc. are portable
Hardware resource on formula electronic equipment is limited and limited battery capacity, so that deep learning text detection is in portable electronic device
The difficulty of upper operation is larger, and user experience is very poor.
The text detection mode that therefore, it is necessary to a kind of speeds of service is fast, low in energy consumption, accuracy rate is high, is particularly suitable for
It is run on portable electronic device, passes through default indicant (such as pointer or hand with Indication Characteristics especially suitable for user
Refer to, specially tip portion or nail) it is directed toward the case where being desired with the region of Text region.
Summary of the invention
The embodiment of the present application provides one kind and is used for character detecting method and device, equipment and computer readable storage medium,
To solve following technical problem in the prior art: existing text detection mode is by factors such as light, noise, background pictures
Influence, the accuracy rate of text detection and identification is very low, or to hardware resource requirements, the high, speed of service is relatively slow, power consumption is high.
The embodiment of the present application adopts the following technical solutions:
A kind of character detecting method, comprising:
Image to be detected of object to be detected is obtained, and image to be detected is converted into the image containing hue information,
In this there is the pixel in the image of hue information to be respectively provided with corresponding tone value；
The tone value of pixel is compared with default tone value range, is obtained by tone value within the scope of default tone value
Pixel composition at least one block of pixels；
Selection meets the block of pixels of preset condition from least one described block of pixels, and the position according to selected block of pixels is true
Determine text detection region.
In the present invention, object to be detected can be any object that user is desired with Text region, can by with
Family is directed toward using default indicant (such as pointer or finger with Indication Characteristics, specially tip portion or nail), so as to
It is accurate to determine the partial region for being desired with Text region.In addition, the resolution ratio of image to be detected can be required here it is lower,
It can even is that the preview graph that portable electronic device is shot by camera, it is therefore desirable to carry out the data volume of image procossing
It is few, be conducive to improve processing speed.Certainly, the image to be detected be also possible to camera shooting picture photo rather than preview
Figure.
Preferably, image to be detected is converted to the image containing hue information can specifically include: will be described to be detected
Image is converted to the image in single channel containing hue information, and image to be detected is preferably converted to the image of HSV format,
Wherein for the pixel in the image of the HSV format other than being respectively provided with tone value, being also respectively provided with indicates color saturation
Intensity value and the brightness value for indicating lightness information.
Preferably, the value range of tone value is 0 to 180, and default tone value may range from 0 to 25.Certainly, here
Default tone value range it is related to target user's hand skin color of detection, can according to the specific colour of skin situation of target user into
The customized setting of row or by the automatic acquistion of Face Detection training, therefore may be other specific ranges, the present invention couple
This is not construed as limiting.
Preferably, the intensity value that the pixel at least one described block of pixels has is being preset within the scope of intensity value,
And the brightness value having is in default brightness value scope.
Preferably, the value range of intensity value is 0 to 255, and default intensity value may range from 58 to 173；It is bright
The value range of angle value is 0 to 255, and default brightness value scope can be 50 to 255.
Preferably, the block of pixels that selection meets preset condition from least one described block of pixels can be with specifically: from institute
It states and selects the maximum block of pixels of area at least one block of pixels.
Preferably, the character detecting method can also include: to be detected from image to be detected comprising default indicant
Indicant image-region, the default indicant preferably with Indication Characteristics pointer, tip portion or nail,
Wherein from least one described block of pixels selection meet preset condition block of pixels specifically may is that from it is described to
The block of pixels containing the indicant image-region is selected in a few block of pixels.
Preferably, the position of the block of pixels according to selected by determines that text detection region can be with specifically: according to selected picture
The position of the indicant image-region in plain block determines the text detection region in image to be detected.
Preferably, from least one described block of pixels selection meet the block of pixels of preset condition after, the method
Can also include: obtained again after user removes default indicant object to be detected, have compared with image to be detected
Another image to be detected of higher resolution, and determine that text detection region can be specific according to the position of selected block of pixels
It is: determines that the region above indicant image-region position is text detection region in described another image to be detected.This hair
In bright, by obtaining the image of object to be detected again after user removes finger, so that there is no user's finger in the image
Possible word segment is covered, the position of predetermined indicant image-region is recycled, can completely identify indicant
The position of image-region and its neighbouring text；The image can use to be detected than determining text detection region before
Image has higher resolution ratio, to improve the accuracy of Text region.
In addition, in the present invention, the region above indicant image-region position refers to that ordinate is greater than instruction in image
The region of the ordinate of object image regional location, the image-region being not limited solely to right above indicant image-region position,
It further include the other image-regions parallel with the image-region right above this, that is to say, that above indicant image-region position
Region include in image all ordinates be greater than indicant image-region position ordinate region.
Preferably, indicant image-region of the detection comprising default indicant may include: from image to be detected
Image to be detected is converted into grayscale image, and using the CascadeClassifier function from OpenCV from
The default indicant image-region is detected in the grayscale image.
In addition, also providing a kind of text detection device corresponding with above-mentioned character detecting method, comprising:
Image processing module, is used to obtain image to be detected of object to be detected, and image to be detected is converted to and is contained
There is the image of hue information, wherein the pixel in the image with hue information is respectively provided with corresponding tone value；
Comparison module is used to for the tone value of pixel being compared with default tone value range, and acquisition is existed by tone value
At least one block of pixels of pixel composition within the scope of default tone value；
Determining module is used to select to meet from least one described block of pixels the block of pixels of preset condition, according to institute
The position of block of pixels is selected to determine text detection region.
Preferably, image to be detected is converted to the image containing hue information can be with specifically: by the mapping to be checked
Picture is converted to the image in single channel containing hue information, and image to be detected is preferably converted to the image of HSV format,
Pixel in the image of middle HSV format is also respectively provided with the saturation for indicating color saturation other than being respectively provided with tone value
Angle value and the brightness value for indicating lightness information.
Preferably, the value range of tone value is 0 to 180, and default tone value range is 0 to 25.
Preferably, the intensity value that the pixel at least one described block of pixels has is being preset within the scope of intensity value,
And the brightness value having is in default brightness value scope.
Preferably, the value range of intensity value is 0 to 255, and default intensity value range is 58 to 173；It is stated clearly
The value range of angle value is 0 to 255, and the default brightness value scope is 50 to 255.
Preferably, the block of pixels that selection meets preset condition from least one described block of pixels can be with specifically: from institute
It states and selects the maximum block of pixels of area at least one block of pixels.
Preferably, image processing module can be also used for indicant of the detection comprising default indicant from image to be detected
Image-region, perform the operation default indicant preferably with Indication Characteristics pointer, tip portion or nail,
The block of pixels that wherein determining module selects to meet from least one described block of pixels preset condition can be with specifically:
Determining module selects the block of pixels containing indicant image-region at least one described block of pixels.
Preferably, determining module can be also used for:
According to the position of indicant image-region described in selected block of pixels, the text detection area in image to be detected is determined
Domain.
Preferably, image processing module can be also used for: select to meet default item from least one described block of pixels
After the block of pixels of part, object to be detected, the tool compared with current image to be detected are obtained again after user removes finger
There is another image to be detected of higher resolution；
And the determining module determines that text detection region may include: described another according to the position of selected block of pixels
Determine that the region above indicant image-region position is text detection region in one image to be detected.
Preferably, described image processing module detects the indicant figure comprising default indicant from described image to be detected
As the step of region may include:
Described image to be detected is converted into grayscale image, and using the CascadeClassifier function of OpenCV from
The default indicant image-region is detected in the grayscale image.
In addition, additionally providing a kind of equipment for being used for text detection accordingly, comprising:
At least one processor；And
The memory being connect at least one described processor communication；Wherein,
Memory is stored with the instruction that can be executed by least one described processor, and described instruction is by described at least one
Device is managed to execute, so that at least one described processor is able to carry out:
Image to be detected of object to be detected is obtained, and image to be detected is converted into the image containing hue information,
In this there is the pixel in the image of hue information to be respectively provided with corresponding tone value；
The tone value of pixel is compared with default tone value range, is obtained by tone value within the scope of default tone value
Pixel composition at least one block of pixels；
Selection meets the block of pixels of preset condition from least one described block of pixels, and the position according to selected block of pixels is true
Determine text detection region.
In addition, additionally providing a kind of non-volatile computer readable storage medium storing program for executing for text detection, it is stored thereon with
Computer executable instructions, the computer executable instructions are set as the execution when being run by processor:
Image to be detected of object to be detected is obtained, and image to be detected is converted into the image containing hue information,
In this there is the pixel in the image of hue information to be respectively provided with corresponding tone value；
The tone value of pixel is compared with default tone value range, is obtained by tone value within the scope of default tone value
Pixel composition at least one block of pixels；
Selection meets the block of pixels of preset condition from least one described block of pixels, and the position according to selected block of pixels is true
Determine text detection region.
At least one above-mentioned technical solution that the embodiment of the present application uses can reach following the utility model has the advantages that by will be to be checked
Altimetric image is converted to such as HSV format of the picture format with hue information, according to hue information and can combine user colour
Setting default tone value range and detect user's hand image-region, so as to quickly and accurately identify text detection
Region, and can be the image that preview graph etc. has low resolution for detecting image to be detected of hand here, it is significant to drop
Image processing data amount before low Text region improves hand and detects speed, fast implements hard recognition and tracking；In addition, knot
Indicant (such as pointer, tip portion or nail with Indication Characteristics) image-region detection algorithm is closed, it can be more accurate
Ground detects the position of user's hand and finger (such as finger tip), avoids the dry of the background color same or similar with user colour
It disturbs, therefore, can prepare to determine that user wishes the character area for detecting and identifying, removal does not need the region of text detection, mentions
High text detection speed reduces equipment power dissipation, improves text detection accuracy, improves user experience.Moreover, determining text inspection
After surveying region, can re-shoot be subsequently used for carrying out text detection, with the image detected for hand compared with higher
The image of resolution ratio, to improve the accuracy of text detection and identification.
Detailed description of the invention
The drawings described herein are used to provide a further understanding of the present application, constitutes part of this application, this Shen
Illustrative embodiments and their description please are not constituted an undue limitation on the present application for explaining the application.In the accompanying drawings:
Fig. 1 is the flow diagram according to the character detecting method of one embodiment of the application；
Fig. 2 is the schematic diagram for being used to carry out image to be detected of hand detection according to an example, which is in practical application
For cromogram；
Image to be detected, is converted into the schematic diagram of HSV format-pattern by the example according to Fig. 3；
The example according to Fig. 4 obtains the schematic diagram of black white image after carrying out threshold transition to HSV image；
Fig. 5 is that the signal of finger nail (i.e. a kind of default indicant) detection is carried out to image to be detected according to another example
Figure, wherein the finger nail that will test is shown with box；
Fig. 6 is to show the schematic diagram of error detection result occur when individually carrying out hand detection using HSV picture format；
Fig. 7 is the schematic diagram according to the text detection device of one embodiment of the application；
Fig. 8 is the schematic diagram according to the text detection equipment of one embodiment of the application.
Specific embodiment
To keep the purposes, technical schemes and advantages of the application clearer, below in conjunction with the application specific embodiment and
Technical scheme is clearly and completely described in corresponding attached drawing.Obviously, described embodiment is only the application one
Section Example, instead of all the embodiments.Based on the embodiment in the application, those of ordinary skill in the art are not doing
Every other embodiment obtained under the premise of creative work out, shall fall in the protection scope of this application.
The scheme of the application is described in detail below.
Fig. 1 is the flow diagram according to the character detecting method of one embodiment of the application.In the process, from equipment angle
For degree, executing subject can be one or more portable electronic device, such as smart phone, intelligent eyes etc., more
It can be to body and the associated functional module of camera in these equipment；For program angle, executing subject correspondingly may be used
It is answered with the program being mounted on these portable electronic devices, such as the application of language translation application, Text region, text casting
With etc..
Process in Fig. 1 may comprise steps of S101-S103.
S101: image to be detected of object to be detected is obtained, and image to be detected is converted into the figure containing hue information
Picture, wherein there is the pixel in the image of hue information to be respectively provided with corresponding tone value for this.
As shown in Figure 2, image to be detected example of acquired object to be detected is shown, image to be detected can lead to
Various modes are crossed to obtain, for example, by camera shoot obtain, can by take pictures obtain or by video recording obtain, certainly to
Detection image can also have other sources, such as from other equipment, or be also possible to ready-made image, the present invention to this not
It is limited.It should be noted that although the picture due to patent application document the substantive requirements of form and be illustrated as grayscale image,
In practical applications, image to be detected should be cromogram.Here object to be detected is that user is desired with Text region
Object can be any object, the invention is not limited in this regard.
In this step, image to be detected can be also converted to containing hue information (such as containing single channel hue information)
Image, pixel therein is respectively provided with corresponding tone value, identifies to be able to use the hue information in subsequent step
The default indicant image section (i.e. block of pixels) of pre-set color is had in the image out, such as is identified with user colour
Finger part, as tip portion or nail or it is other have Indication Characteristics (such as with particular color and/or specific shape)
Any indicant.The example for the image containing hue information that image to be detected that showing in Fig. 3 will acquire is converted to,
Contain hue information in single channel in the example images, this is different from RGB image, and in RGB image, needing will be multiple logical
Road, which is spliced, can just obtain hue information (such as crocus is not belonging to any one in three kinds of Essential colour of RGB, therefore
Need to splice red and green and obtain crocus), thus can not be extracted from image an individual channel and predetermined color into
Row is relatively to identify default indicant.It should be noted that although the Fig. 3 due to patent application document the substantive requirements of form and shown
For grayscale image, but in fact, converting the resulting image containing hue information from image to be detected should also be as cromogram.
S102: the tone value of pixel is compared with default tone value range, is obtained by tone value in default tone value
At least one block of pixels of pixel composition in range.
In this step, it is preassigned according to the color for the indicant for playing indicative function for presetting tone value range.
By the way that the tone value of pixel in image obtained by conversion in step S101 to be compared with default tone value range, this can be derived that
Tone value presets at least one block of pixels composed by the pixel within the scope of tone value at this in image, wherein it is described at least one
Block of pixels includes default indicant image section.That is, by the step special color can be met from screening in image
Adjust the block of pixels of range.Shown in fig. 3 in example, if it is desired to which the Target indication object screened is the finger of people, then will
Default tone value range is set as specific human skin color, then after step S102, can screen in example shown in fig. 3
User's hand of right part in image, except of course that except user's hand, step S102 may also can by book cover with people
Hand color it is identical or approximate image section filters out.As shown in Figure 4, for having hue information shown in Fig. 3
Image carries out hue threshold and compares to be converted into the image after black white image, wherein by its tone value in image obtained by step S101
The color of pixel within the scope of default tone value is changed to white, and tone value is not preset the pixel within the scope of tone value at this
Color be changed to black, it can be seen from the figure that white portion in addition to include user's hand other than, further include other multiple pixels
Block, that is to say, that multiple block of pixels including user's hand have been filtered out in step S101.
S103: selection meets the block of pixels of preset condition from least one described block of pixels, according to selected block of pixels
Position determines text detection region.Herein, preset condition can be varied, is preset.For example, preset condition can
To be the block of pixels comprising preset shape, or the block of pixels of nail etc. comprising identifying.
In the present invention, object to be detected can be any object that user is desired with Text region, can by with
Family is directed toward using default indicant (such as pointer or finger with Indication Characteristics, specially tip portion or nail), so as to
It is accurate to determine the partial region for being desired with Text region.In addition, the resolution ratio of image to be detected can be required here it is lower,
It can even is that the preview graph that portable electronic device is shot by camera, it is therefore desirable to carry out the data volume of image procossing
It is few, be conducive to improve processing speed.Certainly, the image to be detected be also possible to camera shooting picture photo rather than preview
Figure.
Preferably, image to be detected is converted to the image containing hue information can specifically include: by image to be detected
The image of HSV format is converted to, wherein the pixel in the image of the HSV format is also distinguished other than being respectively provided with tone value
Brightness value with the intensity value and expression lightness information that indicate color saturation.HSV (tone H, saturation degree S, lightness V) is
According to the intuitive nature of color by A.R.Smith in a kind of color space of creation in 1978, also referred to as hexagonal pyramid model.
In HSV image, tone value can be measured with angle, and value range is 0 °~360 °, by the inverse time since red
Needle direction calculating, red are 0 °, and green is 120 °, and blue is 240 °, their complementary color is: yellow is 60 °, and cyan is 180 °,
Magenta is 300 °.Certainly, the specific value range of tone value can also carry out linear transformation, to adapt to specifically apply, such as
It is illustrated so that tone value value range is 0 to 180 as an example in the embodiment of the present invention, the invention is not limited in this regard.Saturation degree S
Color is indicated close to the degree of spectrum colour, usual value range is 0%~100%, and value is bigger, and color is more saturated；Certainly, it is saturated
The specific value range of degree can also carry out linear transformation, to adapt to specifically apply, such as with saturation in the embodiment of the present invention
Angle value value range is illustrated for being 0 to 255, the invention is not limited in this regard.Lightness indicates bright degree, right
In light source colour, brightness value is related with the brightness of illuminator；For object color, the transmittance or reflectivity of this value and object have
It closes, usual value range is 0% (black) to 100% (white)；Certainly, the specific value range of lightness can also linearly be become
It changes, is illustrated so that brightness value value range is 0 to 255 as an example to adapt to specifically apply, such as in the embodiment of the present invention, this
Invention to this with no restriction.
It will be appreciated by those skilled in the art that the three dimensional representation of HSV iconic model can develop from RGB Cube,
It is the examples from RGB image format to HSV image format conversion below such as:
R '=R/255
G '=G/255
B '=B/255
C max=max (R, G, B)
C min=min (R ', G ', B ')
Δ=C max-C min
H is calculated:
S is calculated:
V is calculated:
V=C max
Wherein R, G, B indicate red, the green and blue value of each pixel in RGB image format, and H, S and V indicate HSV figure
The tone value of each pixel, intensity value and brightness value in format, value range is respectively 0-360 °, 0-1 and 0-1.
In the present embodiment, the value range of tone value is 0 to 180, and default tone value may range from 0 to 25.When
So, here default tone value range is related to target user's hand skin color of detection, can be according to the specific skin of target user
Pornographic condition carries out customized setting or by the automatic acquistion of Face Detection training, therefore may be other specific ranges,
This is not limited by the present invention.
In addition, when the tone value of each pixel is compared with default tone value range in HSV image, can also will
Its corresponding protection scope is compared the intensity value and brightness value of pixel respectively, selects tone value, intensity value and bright
The angle value block of pixels that the pixel in corresponding range forms respectively, so that the pixel at least one described block of pixels had
Intensity value is within the scope of default intensity value, and the brightness value having is in default brightness value scope.
For example, the value range of intensity value is 0 to 255, and default intensity value may range from 58 to 173；Lightness
The value range of value is 0 to 255, and default brightness value scope can be 50 to 255.
In the present embodiment, at least one obtained described block of pixels might not be entirely desired block of pixels, example
It such as might not be entirely the block of pixels comprising default indicant, it is therefore necessary to further be screened to it.For example, from institute
State at least one block of pixels selection meet the block of pixels of preset condition can be with specifically: at least one described block of pixels
Select the maximum block of pixels of area.
In addition, presence and its position by detecting the default indicant in image, are conducive to more accurately detect
Target text detection zone out.For example, the character detecting method can also include: to detect from image to be detected comprising default
The indicant image-region of indicant, the default indicant preferably with Indication Characteristics pointer, tip portion or nail,
But it is also possible to other objects with default specific shape.When detecting default indicant (such as nail), can use
The existing available algorithm from OpenCV is realized, for example, by using the CascadeClassifier function in OpenCV
(https: //docs.opencv.org/3.2.0/d1/de5/classcv_1_1CascadeClassifie r.html) come real
It is existing.Before using the function, needs image to be detected being converted into grayscale image, which is fed as input to the letter
Number, and those skilled in the art will recognize that remaining the input parameter for according to specific application and setting the function, such as
ScaleFactor (input picture carries out reduced scale, such as value 1.1 in identification every time), minNeighbors are (i.e.
Each frame, which needs minimum how many to close on frame, can just determine that it is believable frame, such as value 3), the value of minimum frame (such as value
24 × 24 block of pixels), the value (such as 50 × 50 block of pixels of value) of maximum frame, be enable to detect from the grayscale image
Default indicant image-region, wherein any existing mode, this hair can be used when image to be detected is converted into grayscale image
It is bright that this is also not limited.
Using the function, by study, can be recognized accurately pointer in image with Indication Characteristics, tip portion or
The indicants such as nail portions, so that the present invention can be more quasi- when user indicates target text region by the indicant
Really according to the position of the indicant identified (such as shown in label 500 in Fig. 5), determine in image pointed by the indicant
Accurate word segment (i.e. target text detection zone), and unnecessary Text region region is eliminated, so that subsequent
Text region is more efficient.
On the contrary, it can be seen that carrying out at least one described block of pixels into one in conjunction with default indicant from the example of Fig. 6
Step screening, can be further improved the accuracy of text detection region recognition.As shown in fig. 6, the pixel determined according to tone value
Block will include the cabinet image section 600 in the lower right corner in hand images part and image in figure (because of cabinet color and hand
Color is close), and since cabinet seems bigger than hand in the picture, so when being selected according to maximum pixel block principle, it can
Word segment indicated by cabinet image section can be determined as target text detection zone to identify, so default instruction
Object can be conducive to more accurately identify text detection region.
In another example, from least one described block of pixels selection meet the block of pixels of preset condition after, institute
The method of stating can also include: obtained again after user removes default indicant object to be detected, with image to be detected phase
Determine that text detection region can be with than another image to be detected with higher resolution, and according to the position of selected block of pixels
Specifically: determining that the region above indicant image-region position is text detection region in described another image to be detected.
In the present invention, by obtaining the image of object to be detected again after user removes finger, so that there is no user in the image
Finger covers possible word segment, recycles the position of predetermined indicant image-region, can completely identify and refer to
Show object image region position and its neighbouring text；The image can than before for determine text detection region and use to
Detection image has higher resolution ratio, to improve the accuracy of Text region.
In addition, in the present invention, the region above indicant image-region position refers to that ordinate is greater than instruction in image
The region of the ordinate of object image regional location, the image-region being not limited solely to right above indicant image-region position,
It further include the other image-regions parallel with the image-region right above this, that is to say, that above indicant image-region position
Region include in image all ordinates be greater than indicant image-region position ordinate region.
It should be noted that above-named some specific examples reflect the thought of application scheme, but it is not
Unique form of implementation, the thought based on application scheme can obtain more similar examples.
Based on same thinking, some embodiments of the present application additionally provide the corresponding text detection device of the above method,
Equipment and nonvolatile computer storage media.
As shown in fig. 7, showing the schematic diagram of the text detection device according to one embodiment of the application, which includes:
Image processing module 701 is used to obtain image to be detected of object to be detected, and image to be detected is converted to
Image containing hue information, wherein the pixel in the image with hue information is respectively provided with corresponding tone value；
Comparison module 702 is used to for the tone value of pixel being compared with default tone value range, obtains by tone value
At least one block of pixels of pixel composition within the scope of default tone value；
Determining module 703 is used to select to meet from least one described block of pixels the block of pixels of preset condition, according to
The position of selected block of pixels determines text detection region.
Image to be detected is converted to the image containing hue information can be with specifically: image to be detected is converted to HSV
The image of format, wherein the pixel in the image of HSV format is also respectively provided with expression color other than being respectively provided with tone value
The intensity value of saturation degree and the brightness value for indicating lightness information.
In one example, the value range of tone value is 0 to 180, and default tone value range is 0 to 25.
The intensity value that pixel at least one described block of pixels has is being preset within the scope of intensity value, and have
Brightness value is in default brightness value scope.
In one example, the value range of intensity value is 0 to 255, and default intensity value range is 58 to 173；Institute
The value range for stating angle value clearly is 0 to 255, and the default brightness value scope is 50 to 255.
The block of pixels that selection meets preset condition from least one described block of pixels can be with specifically: from described at least one
The maximum block of pixels of area is selected in a block of pixels.
Image processing module 701 can be also used for instruction object image of the detection comprising default indicant from image to be detected
Region, the default indicant can be pointer, tip portion or nail with Indication Characteristics,
Wherein determining module 703 selects the block of pixels for meeting preset condition can be specific from least one described block of pixels
Are as follows: determining module 703 is selected at least one described block of pixels containing detected default indicant image-region
Block of pixels.
Determining module 703 can be also used for: the position of indicant image-region be preset according to selected block of pixels, really
Determine the text detection region in image to be detected.
Image processing module 701 can be also used for: select to meet preset condition from least one described block of pixels
After block of pixels, obtained again after user removes finger object to be detected, compared with current image to be detected have more
High-resolution another image to be detected；
And determining module 703 determines that text detection region may include: described another according to the position of selected block of pixels
Determine that the region above default indicant image-region position is text detection region in one image to be detected.
In addition, image processing module 701 detects the indicant image-region comprising default indicant from image to be detected
The step of may include:
Image to be detected is converted into grayscale image, and uses the CascadeClassifier function of OpenCV from the ash
Default indicant image-region is detected in degree figure.
At least one above-mentioned technical solution that the embodiment of the present application uses can reach following the utility model has the advantages that by will be to be checked
Altimetric image is converted to such as HSV format of the picture format with hue information, according to hue information and can combine user colour
Setting default tone value range and detect user's hand image-region, so as to quickly and accurately identify text detection
Region, and can be the image that preview graph etc. has low resolution for detecting image to be detected of hand here, it is significant to drop
Image processing data amount before low Text region improves hand and detects speed, fast implements hard recognition and tracking；In addition, knot
Indicant image-region detection algorithm is closed, can detect the position of user's hand and finger (such as finger tip) more accurately,
The interference of the background color same or similar with user colour is avoided, therefore, can prepare to determine that user wishes to detect and identify
Character area, removal do not need the region of text detection, improves text detection speed, reduce equipment power dissipation, improve text inspection
Accuracy is surveyed, user experience is improved.Moreover, can re-shoot after determining text detection region and be subsequently used for carrying out text
Detection, with the image detected for hand compared with higher resolution image, thus improve text detection and identify
Accuracy.
Fig. 8 is according to the schematic diagram of the text detection equipment of one embodiment of the application, which includes:
At least one processor；And
The memory being connect at least one described processor communication；Wherein,
Memory is stored with the instruction that can be executed by least one described processor, and described instruction is by described at least one
Device is managed to execute, so that at least one described processor is able to carry out:
Image to be detected of object to be detected is obtained, and image to be detected is converted into the image containing hue information,
In this there is the pixel in the image of hue information to be respectively provided with corresponding tone value；
The tone value of pixel is compared with default tone value range, is obtained by tone value within the scope of default tone value
Pixel composition at least one block of pixels；
Selection meets the block of pixels of preset condition from least one described block of pixels, and the position according to selected block of pixels is true
Determine text detection region.
According to some embodiments of the present application, provide corresponding with character detecting method described above for text detection
Nonvolatile computer storage media, be stored thereon with computer executable instructions, which is set as
The execution when being run by processor:
Image to be detected of object to be detected is obtained, and image to be detected is converted into the image containing hue information,
In this there is the pixel in the image of hue information to be respectively provided with corresponding tone value；
The tone value of pixel is compared with default tone value range, is obtained by tone value within the scope of default tone value
Pixel composition at least one block of pixels；
Selection meets the block of pixels of preset condition from least one described block of pixels, and the position according to selected block of pixels is true
Determine text detection region.
Various embodiments are described in a progressive manner in the application, same and similar part between each embodiment
It may refer to each other, each embodiment focuses on the differences from other embodiments.Especially for device, set
For standby and computer readable storage medium embodiment, since it is substantially similar to the method embodiment, so its description carries out
Simplify, related place can be found in the part explanation of embodiment of the method.
Device, equipment and computer readable storage medium provided by the embodiments of the present application and method be it is one-to-one, because
This, device, equipment and computer readable storage medium also have the similar advantageous effects of corresponding method, due to upper
Face is described in detail the advantageous effects of method, and therefore, which is not described herein again, and device, equipment and computer can
Read the advantageous effects of storage medium.
It should be understood by those skilled in the art that, the embodiment of the present invention can provide as method, system or computer program
Product.Therefore, complete hardware embodiment, complete software embodiment or reality combining software and hardware aspects can be used in the present invention
Apply the form of example.Moreover, it wherein includes the computer of computer usable program code that the present invention, which can be used in one or more,
The computer program implemented in usable storage medium (including but not limited to magnetic disk storage, CD-ROM, optical memory etc.) produces
The form of product.
The present invention be referring to according to the method for the embodiment of the present invention, the process of equipment (system) and computer program product
Figure and/or block diagram describe.It should be understood that every one stream in flowchart and/or the block diagram can be realized by computer program instructions
The combination of process and/or box in journey and/or box and flowchart and/or the block diagram.It can provide these computer programs
Instruct the processor of general purpose computer, special purpose computer, Embedded Processor or other programmable data processing devices to produce
A raw machine, so that being generated by the instruction that computer or the processor of other programmable data processing devices execute for real
The device for the function of being specified in present one or more flows of the flowchart and/or one or more blocks of the block diagram.
These computer program instructions, which may also be stored in, is able to guide computer or other programmable data processing devices with spy
Determine in the computer-readable memory that mode works, so that it includes referring to that instruction stored in the computer readable memory, which generates,
Enable the manufacture of device, the command device realize in one box of one or more flows of the flowchart and/or block diagram or
The function of being specified in multiple boxes.
These computer program instructions also can be loaded onto a computer or other programmable data processing device, so that counting
Series of operation steps are executed on calculation machine or other programmable devices to generate computer implemented processing, thus in computer or
The instruction executed on other programmable devices is provided for realizing in one or more flows of the flowchart and/or block diagram one
The step of function of being specified in a box or multiple boxes.
In a typical configuration, calculating equipment includes one or more processors (CPU), input/output interface, net
Network interface and memory.
Memory may include the non-volatile memory in computer-readable medium, random access memory (RAM) and/or
The forms such as Nonvolatile memory, such as read-only memory (ROM) or flash memory (flash RAM).Memory is computer-readable medium
Example.
Computer-readable medium includes permanent and non-permanent, removable and non-removable media can be by any method
Or technology come realize information store.Information can be computer readable instructions, data structure, the module of program or other data.
The example of the storage medium of computer includes, but are not limited to phase change memory (PRAM), static random access memory (SRAM), moves
State random access memory (DRAM), other kinds of random access memory (RAM), read-only memory (ROM), electric erasable
Programmable read only memory (EEPROM), flash memory or other memory techniques, read-only disc read only memory (CD-ROM) (CD-ROM),
Digital versatile disc (DVD) or other optical storage, magnetic cassettes, tape magnetic disk storage or other magnetic storage devices
Or any other non-transmission medium, can be used for storage can be accessed by a computing device information.
It should also be noted that, the terms "include", "comprise" or its any other variant are intended to nonexcludability
It include so that the process, method, commodity or the equipment that include a series of elements not only include those elements, but also to wrap
Include other elements that are not explicitly listed, or further include for this process, method, commodity or equipment intrinsic want
Element.In the absence of more restrictions, the element limited by sentence "including a ...", it is not excluded that including described want
There is also other identical elements in the process, method of element, commodity or equipment.
The above description is only an example of the present application, is not intended to limit this application.For those skilled in the art
For, various changes and changes are possible in this application.All any modifications made within the spirit and principles of the present application are equal
Replacement, improvement etc., should be included within the scope of the claims of this application.
Claims (22)
1. a kind of character detecting method characterized by comprising
Image to be detected of object to be detected is obtained, and image to be detected is converted into the image containing hue information, wherein institute
It states the pixel in the image with hue information and is respectively provided with corresponding tone value；
The tone value of the pixel is compared with default tone value range, is obtained by tone value in the default tone value model
At least one block of pixels of pixel composition in enclosing；
Selection meets the block of pixels of preset condition from least one described block of pixels, determines text according to the position of selected block of pixels
Word detection zone.
2. character detecting method as described in claim 1, which is characterized in that described to be converted to image to be detected containing tone
The image of information includes: that described image to be detected is converted to image in single channel containing hue information, preferably by institute
The image that image to be detected is converted to HSV format is stated, wherein the pixel in the image of the HSV format is described in addition to being respectively provided with
Except tone value, also it is respectively provided with the intensity value for indicating color saturation and indicates the brightness value of lightness information.
3. character detecting method as claimed in claim 2, which is characterized in that the value range of the tone value is 0 to 180,
And the default tone value range is 0 to 25.
4. character detecting method as claimed in any one of claims 1-3, which is characterized in that at least one described block of pixels
The intensity value that has of pixel within the scope of default intensity value, and the brightness value having is in default brightness value scope.
5. character detecting method as claimed in claim 4, which is characterized in that the value range of the intensity value be 0 to
255, and the default intensity value range is 58 to 173；The value range of the brightness value be 0 to 255, and it is described preset it is bright
Angle value range is 50 to 255.
6. character detecting method according to any one of claims 1 to 5, which is characterized in that from least one described block of pixels
The block of pixels that middle selection meets preset condition includes: the selection maximum block of pixels of area at least one described block of pixels.
7. character detecting method according to any one of claims 1 to 5, which is characterized in that further include: from described to be detected
Indicant image-region of the detection comprising default indicant, finger of the default indicant preferably with Indication Characteristics in image
Needle, tip portion or nail,
It includes: from least one described pixel that wherein selection, which meets the block of pixels of preset condition, from least one described block of pixels
The block of pixels for containing the indicant image-region is selected in block.
8. character detecting method as claimed in claim 7, which is characterized in that the position of the block of pixels according to selected by determines text
Word detection zone includes:
According to the position of indicant image-region described in selected block of pixels, the text detection region in image to be detected is determined.
9. character detecting method as claimed in claim 7, which is characterized in that in the selector from least one described block of pixels
After the block of pixels for closing preset condition, the method also includes:
Obtained again after user removes the default indicant object to be detected, compared with described image to be detected
Another image to be detected with higher resolution, and the position of the block of pixels according to selected by determines that text detection region is wrapped
It includes: determining that the region above indicant image-region position is text detection region in described another image to be detected.
10. character detecting method as claimed in any one of claims 7-9, which is characterized in that from described image to be detected
Detection includes that the indicant image-region of default indicant includes:
Described image to be detected is converted into grayscale image, and using CascadeClassifier function from the grayscale image
Detect the default indicant image-region.
11. a kind of text detection device characterized by comprising
Image processing module is used to obtain image to be detected of object to be detected, and image to be detected is converted to containing coloured
The image of information is adjusted, wherein the pixel in the image with hue information is respectively provided with corresponding tone value；
Comparison module is used to for the tone value of the pixel being compared with default tone value range, and acquisition is existed by tone value
At least one block of pixels of pixel composition within the scope of the default tone value；
Determining module is used to select to meet from least one described block of pixels the block of pixels of preset condition, according to selected picture
The position of plain block determines text detection region.
12. text detection device as claimed in claim 11, which is characterized in that described to be converted to image to be detected containing coloured
The image for adjusting information includes: that described image to be detected is converted to image in single channel containing hue information, preferably will
Described image to be detected is converted to the image of HSV format, wherein the pixel in the image of the HSV format is in addition to being respectively provided with
It states except tone value, be also respectively provided with the intensity value for indicating color saturation and indicate the brightness value of lightness information.
13. text detection device as claimed in claim 12, which is characterized in that the value range of the tone value be 0 to
180, and the default tone value range is 0 to 25.
14. the text detection device as described in any one of claim 11-13, which is characterized in that at least one described pixel
The intensity value that pixel in block has is within the scope of default intensity value, and the brightness value having is in default brightness value scope
It is interior.
15. text detection device as claimed in claim 14, which is characterized in that the value range of the intensity value be 0 to
255, and the default intensity value range is 58 to 173；The value range of the brightness value be 0 to 255, and it is described preset it is bright
Angle value range is 50 to 255.
16. the text detection device as described in any one of claim 11-15, which is characterized in that from least one described picture
It includes: the selection maximum pixel of area at least one described block of pixels that selection, which meets the block of pixels of preset condition, in plain block
Block.
17. the text detection device as described in any one of claim 11-15, which is characterized in that described image processing module
It is also used to detect the indicant image-region comprising the default indicant, the default indicant from described image to be detected
Preferably with the pointer, tip portion or nail of Indication Characteristics,
It includes: described true that wherein the determining module selects the block of pixels for meeting preset condition from least one described block of pixels
Cover half block selects the block of pixels containing the indicant image-region at least one described block of pixels.
18. text detection device as claimed in claim 17, which is characterized in that the determining module is used for:
According to the position of indicant image-region described in selected block of pixels, the text detection region in image to be detected is determined.
19. text detection device as claimed in claim 17, which is characterized in that described image processing module is also used to: from
After selection meets the block of pixels of preset condition at least one described block of pixels, after user removes the default indicant
Obtain again the object to be detected, compared with described image to be detected with higher resolution another image to be detected；
And the determining module determines that text detection region includes: described another to be detected according to the position of selected block of pixels
Determine that the region above indicant image-region position is text detection region in image.
20. the text detection device as described in any one of claim 17-19, which is characterized in that described image processing module
Indicant image-region of the detection comprising default indicant includes: from described image to be detected
Described image to be detected is converted into grayscale image, and using CascadeClassifier function from the grayscale image
Detect the default indicant image-region.
21. a kind of equipment for text detection characterized by comprising
At least one processor；And
The memory being connect at least one described processor communication；Wherein,
The memory is stored with the instruction that can be executed by least one described processor, and described instruction is by described at least one
Device is managed to execute, so that at least one described processor is able to carry out:
Image to be detected of object to be detected is obtained, and image to be detected is converted into the image containing hue information, wherein institute
It states the pixel in the image with hue information and is respectively provided with corresponding tone value；
The tone value of the pixel is compared with default tone value range, is obtained by tone value in the default tone value model
At least one block of pixels of pixel composition in enclosing；
Selection meets the block of pixels of preset condition from least one described block of pixels, determines text according to the position of selected block of pixels
Word detection zone.
22. a kind of non-volatile computer readable storage medium storing program for executing for text detection is stored thereon with the executable finger of computer
It enables, which is characterized in that the computer executable instructions are set as the execution when being run by processor:
Image to be detected of object to be detected is obtained, and image to be detected is converted into the image containing hue information, wherein institute
It states the pixel in the image with hue information and is respectively provided with corresponding tone value；
The tone value of the pixel is compared with default tone value range, is obtained by tone value in the default tone value model
At least one block of pixels of pixel composition in enclosing；
Selection meets the block of pixels of preset condition from least one described block of pixels, determines text according to the position of selected block of pixels
Word detection zone.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910184219.4A CN110070080A (en) | 2019-03-12 | 2019-03-12 | A kind of character detecting method and device, equipment and computer readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910184219.4A CN110070080A (en) | 2019-03-12 | 2019-03-12 | A kind of character detecting method and device, equipment and computer readable storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110070080A true CN110070080A (en) | 2019-07-30 |
Family
ID=67365189
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910184219.4A Pending CN110070080A (en) | 2019-03-12 | 2019-03-12 | A kind of character detecting method and device, equipment and computer readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110070080A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111383054A (en) * | 2020-03-10 | 2020-07-07 | 中国联合网络通信集团有限公司 | Advertisement checking method and device |
| CN111445466A (en) * | 2020-04-01 | 2020-07-24 | 济南浪潮高新科技投资发展有限公司 | Bolt anti-screwing-leakage detection method and device and medium |
| CN111711758A (en) * | 2020-06-29 | 2020-09-25 | 广东小天才科技有限公司 | Multi-pointing test question shooting method and device, electronic equipment and storage medium |
| CN112580594A (en) * | 2020-12-29 | 2021-03-30 | 浙江大华技术股份有限公司 | Document identification method and device, computer equipment and storage medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102169540A (en) * | 2011-03-28 | 2011-08-31 | 汉王科技股份有限公司 | Camera-based point reading positioning method and device |
| CN107424125A (en) * | 2017-04-14 | 2017-12-01 | 深圳市金立通信设备有限公司 | A kind of image weakening method and terminal |
| CN108875694A (en) * | 2018-07-04 | 2018-11-23 | 百度在线网络技术（北京）有限公司 | Speech output method and device |
| CN109325464A (en) * | 2018-10-16 | 2019-02-12 | 上海翎腾智能科技有限公司 | A kind of finger point reading character recognition method and interpretation method based on artificial intelligence |
-
2019
- 2019-03-12 CN CN201910184219.4A patent/CN110070080A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102169540A (en) * | 2011-03-28 | 2011-08-31 | 汉王科技股份有限公司 | Camera-based point reading positioning method and device |
| CN107424125A (en) * | 2017-04-14 | 2017-12-01 | 深圳市金立通信设备有限公司 | A kind of image weakening method and terminal |
| CN108875694A (en) * | 2018-07-04 | 2018-11-23 | 百度在线网络技术（北京）有限公司 | Speech output method and device |
| CN109325464A (en) * | 2018-10-16 | 2019-02-12 | 上海翎腾智能科技有限公司 | A kind of finger point reading character recognition method and interpretation method based on artificial intelligence |
Non-Patent Citations (3)
| Title |
|---|
| RAINBOW想要OFFER: ""opencv的cascade分类器源码分析"", 《CSDN》 * |
| 廖宁放等: "《数字图文图像颜色管理系统概论》", 31 May 2009, 北京理工大学出版社 * |
| 高志强等: "《深度学习：从入门到实战》", 30 June 2018, 中国铁道出版社 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111383054A (en) * | 2020-03-10 | 2020-07-07 | 中国联合网络通信集团有限公司 | Advertisement checking method and device |
| CN111445466A (en) * | 2020-04-01 | 2020-07-24 | 济南浪潮高新科技投资发展有限公司 | Bolt anti-screwing-leakage detection method and device and medium |
| CN111445466B (en) * | 2020-04-01 | 2023-05-05 | 山东浪潮科学研究院有限公司 | Bolt anti-leakage screwing detection method, equipment and medium |
| CN111711758A (en) * | 2020-06-29 | 2020-09-25 | 广东小天才科技有限公司 | Multi-pointing test question shooting method and device, electronic equipment and storage medium |
| CN111711758B (en) * | 2020-06-29 | 2021-06-18 | 广东小天才科技有限公司 | Multi-pointing test question shooting method and device, electronic equipment and storage medium |
| CN112580594A (en) * | 2020-12-29 | 2021-03-30 | 浙江大华技术股份有限公司 | Document identification method and device, computer equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110070080A (en) | A kind of character detecting method and device, equipment and computer readable storage medium | |
| US11457138B2 (en) | Method and device for image processing, method for training object detection model | |
| US9292911B2 (en) | Automatic image adjustment parameter correction | |
| CN103646392B (en) | Backlighting detecting and equipment | |
| CN108322646A (en) | Image processing method, device, storage medium and electronic equipment | |
| CN102576456B (en) | The luminance digital image of use range information is adjusted | |
| CN110473185A (en) | Image processing method and device, electronic equipment, computer readable storage medium | |
| CN103617432A (en) | Method and device for recognizing scenes | |
| CN110191291A (en) | Image processing method and device based on multiple image | |
| CN110290289A (en) | Image denoising method, device, electronic equipment and storage medium | |
| CN104182721A (en) | Image processing system and image processing method capable of improving face identification rate | |
| JP2007047965A (en) | Method and device for detecting object of digital image, and program | |
| CN107103606A (en) | A kind of image-recognizing method and device | |
| CN103098078A (en) | Smile detection systems and methods | |
| JP2013140428A (en) | Edge detection device, edge detection program, and edge detection method | |
| CN108322651A (en) | Image pickup method and device, electronic equipment, computer readable storage medium | |
| CN109005369A (en) | Exposal control method, device, electronic equipment and computer readable storage medium | |
| CN108564558A (en) | Wide dynamic images processing method, device, equipment and storage medium | |
| Yan et al. | Optimization research on deep learning and temporal segmentation algorithm of video shot in basketball games | |
| CN102333174A (en) | Video image processing method and device for the same | |
| US20160140748A1 (en) | Automated animation for presentation of images | |
| US20190205689A1 (en) | Method and device for processing image, electronic device and medium | |
| CN109191398A (en) | Image processing method, device, computer readable storage medium and electronic equipment | |
| CN102447913B (en) | Color migration processing method and system | |
| KR20180080618A (en) | Method and apparatus for realistic rendering based augmented reality |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |