CN109146834A - Car damage identification method and device, computer readable storage medium, terminal - Google Patents
Car damage identification method and device, computer readable storage medium, terminal Download PDFInfo
- Publication number
- CN109146834A CN109146834A CN201710443445.0A CN201710443445A CN109146834A CN 109146834 A CN109146834 A CN 109146834A CN 201710443445 A CN201710443445 A CN 201710443445A CN 109146834 A CN109146834 A CN 109146834A
- Authority
- CN
- China
- Prior art keywords
- picture
- damage
- difference
- setting loss
- submodule
- 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
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
- G06T7/001—Industrial image inspection using an image reference approach
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/12—Edge-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/187—Segmentation; Edge detection involving region growing; involving region merging; involving connected component labelling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/20—Administration of product repair or maintenance
-
- 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
- G06T2207/10—Image acquisition modality
- G06T2207/10004—Still image; Photographic image
-
- 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
- G06T2207/10—Image acquisition modality
- G06T2207/10024—Color image
Abstract
A kind of car damage identification method and device, computer readable storage medium, terminal, the described method comprises the following steps: determine setting loss picture, the setting loss picture is shown with damage component;Determine the difference of the setting loss picture and template picture, the template picture is shown with the normal component with the damage parts match;Damaged condition is determined according to the difference.The present invention program can accurately determine damaged condition, improve convenience by the photo for the position of collision or damage component that user provides.
Description
Technical field
The present invention relates to automobile technical field, more particularly, to a kind of car damage identification method and device, computer-readable deposit
Storage media, terminal.
Background technique
In the prior art, when vehicle collision or the damage of other forms occurs, user or setting loss person is needed to arrive in person
Scene confirmation damaged condition, efficiency is lower, and convenience is poor.
Although some users can take pictures to position of collision or damage component, and then prejudged and damaged according to setting loss photo
Situation.But for the photo of position of collision or damage component, user or setting loss person also can only observe by the naked eye and rely on
Micro-judgment damaged condition, accuracy rate are poor.
Summary of the invention
The technical problem to be solved by the present invention is to provide a kind of car damage identification method and device, computer readable storage medium,
Terminal can accurately determine damaged condition, improve convenience by the photo for the position of collision or damage component that user provides
Property.
In order to solve the above technical problems, the embodiment of the present invention provides a kind of car damage identification method, comprising the following steps: determine
Setting loss picture, the setting loss picture is shown with damage component;Determine the difference of the setting loss picture and template picture, the template
Picture is shown with the normal component with the damage parts match;Damaged condition is determined according to the difference.
Optionally, the determining setting loss picture includes: the damage information for obtaining user and uploading, and the damage information includes showing
The uploading pictures of the damage component are gone out;The setting loss picture is extracted from the uploading pictures.
Optionally, the damage information further includes position of the damage component on vehicle.
It optionally, include: that institute is identified from the uploading pictures from the setting loss picture is extracted in the uploading pictures
State damage component;According to the damage component identified, the setting loss picture is extracted from the uploading pictures.
Optionally, identify that the damage component includes: using image recognition algorithm from described from the uploading pictures
The damage component is identified in uploading pictures.
Optionally, from extracted in the uploading pictures setting loss picture include: using image segmentation algorithm from described
The setting loss picture is extracted in blit piece.
Optionally, described image partitioning algorithm includes: the image segmentation algorithm based on edge detection, the image based on threshold value
Partitioning algorithm or image segmentation algorithm based on region growing.
Optionally, the difference for determining the setting loss picture and template picture includes: to determine the setting loss using hash algorithm
The difference of picture and template picture.
Optionally, the difference that the setting loss picture and template picture are determined using hash algorithm includes: the use
Hash algorithm determines that the difference of the setting loss picture and template picture includes: cryptographic Hash and the mould based on the setting loss picture
The cryptographic Hash of plate picture determines the difference.
Optionally, constructing cryptographic Hash for the setting loss picture includes: the ash that the setting loss picture is converted to pre-set level
Picture is spent, and is divided into present count destination, there are multiple pixels in each cell, each pixel has ash
Angle value;In each cell, the average value of the gray value of the multiple pixel is calculated, using the gray scale as each cell
Average value;Based on the average gray of each cell, total average gray of the present count destination is calculated;
Each cell of the gray scale picture is traversed, if the average gray of the cell is flat more than or equal to total gray scale
The record result of mean value, the cell is the first numerical value, is otherwise the second value different from first numerical value;It will be described
For the record result of present count destination as cryptographic Hash, the digit of the cryptographic Hash is identical as the preset number.
Optionally, the setting loss picture is being switched to the gray scale picture of pre-set level, and is being divided into the list of preset number
Before first lattice, cryptographic Hash is constructed for the setting loss picture further include: reduce the size of the setting loss picture.
Optionally, the cryptographic Hash of cryptographic Hash and the template picture based on the setting loss picture, determines the difference packet
Include: the record of setting loss picture described in successive appraximation and the template picture is as a result, calculate the number of the different cell of record result
Mesh, using as the difference.
Optionally, determine that damaged condition includes: according to the difference and one or more preset thresholds according to the difference
Comparison result, judge damage grade.
Optionally, according to the comparison result of the difference and one or more preset thresholds, judge to damage grade include: as
Difference described in fruit then judges the damage grade for no damage less than the first preset threshold;If the difference is more than or equal to the
One preset threshold and less than the second preset threshold then judges the damage grade for slight damage;If the difference be greater than etc.
In the second preset threshold and it is less than third predetermined threshold value, then judges the damage grade for moderate damage;If the difference is big
In being equal to third predetermined threshold value and less than the 4th preset threshold, then judge that the damage grade is damaged for severe;If the difference
Value is more than or equal to the 4th preset threshold, then is judged as null result;Wherein, it is default to be less than described second for first preset threshold
Threshold value, second preset threshold are less than the third predetermined threshold value, and the third predetermined threshold value is less than the 4th default threshold
Value.
Optionally, the car damage identification method further include: for the damage component, query suggestion buy information and/
Or suggest repair message;It issues the user with the suggestion purchase information and/or suggests repair message.
In order to solve the above technical problems, the embodiment of the present invention provides a kind of car damage identification device, comprising: picture determines mould
Block is adapted to determine that setting loss picture, and the setting loss picture is shown with damage component;Difference determining module is adapted to determine that the setting loss
The difference of picture and template picture, the template picture is shown with the normal component with the damage parts match;Damage determines
Module, suitable for determining damaged condition according to the difference.
In order to solve the above technical problems, the embodiment of the present invention provides a kind of computer readable storage medium, it is stored thereon with
The step of computer instruction, the computer instruction executes above-mentioned car damage identification method when running.
In order to solve the above technical problems, the embodiment of the present invention provides a kind of terminal, including memory and processor, it is described to deposit
The computer instruction that can be run on the processor is stored on reservoir, when the processor runs the computer instruction
The step of executing above-mentioned car damage identification method.
Compared with prior art, the technical solution of the embodiment of the present invention has the advantages that
In embodiments of the present invention, setting loss picture is determined, the setting loss picture is shown with damage component;Determine the setting loss
The difference of picture and template picture, the template picture is shown with the normal component with the damage parts match;According to described
Difference determines damaged condition.It using the above scheme, can be quasi- by the photo for the position of collision or damage component that user provides
It really determines the difference of setting loss picture and template picture, to accurately determine damaged condition, improves convenience.
Further, in embodiments of the present invention, for damaging component, inquire and provide purchase refill-unit and/or maintenance damage
The suggestion of bad component enhances convenience and economy.
Detailed description of the invention
Fig. 1 is a kind of flow chart of car damage identification method in the embodiment of the present invention;
Fig. 2 is a kind of flow chart of specific implementation of step S11 in Fig. 1;
Fig. 3 is a kind of flow chart of specific implementation of step S22 in Fig. 2;
The step of Fig. 4 is a kind of difference that setting loss picture and template picture are determined using hash algorithm in the embodiment of the present invention
Flow chart;
Fig. 5 is a kind of flow chart of specific embodiment of step S41 in Fig. 4;
Fig. 6 is a kind of comparison result according to difference and one or more preset thresholds in the embodiment of the present invention, judges to damage
The flow chart of the step of bad grade;
Fig. 7 is a kind of structural schematic diagram of car damage identification device in the embodiment of the present invention;
Fig. 8 is a kind of structural schematic diagram of specific embodiment of picture determining module 71 in Fig. 7;
Fig. 9 is a kind of structural schematic diagram of specific embodiment of extracting sub-module 82 in Fig. 8;
Figure 10 is the structural schematic diagram that a kind of difference determines submodule in the embodiment of the present invention;
Figure 11 is a kind of structural schematic diagram of specific embodiment of cryptographic Hash construction submodule 101 in Figure 10;
Figure 12 is a kind of structural schematic diagram of grade judging submodule in the embodiment of the present invention.
Specific embodiment
In the prior art, when vehicle collision or the damage of other forms occurs, user or setting loss person is needed to arrive in person
Scene confirmation damaged condition, efficiency is lower, and convenience is poor.Although some users can be to position of collision or damage component
It takes pictures, and then damaged condition is prejudged according to scene photograph.But it for the photo of position of collision or damage component, uses
Family or setting loss person also can only observe by the naked eye and rely on micro-judgment damaged condition, and accuracy rate is poor.
The present inventor has found that the prior art relies primarily on the manual operation of user or setting loss person, nothing after study
Method accurately judges damaged condition by setting loss photo.This is because different user or different setting loss persons sentence with different
Disconnected experience and setting loss skill, no matter in terms of the accuracy of judgement or the accuracy of setting loss, manual operation is all extremely difficult to machine
The level that tool is automatically brought into operation is easy to happen error in judgement or the improper problem of setting loss.
In embodiments of the present invention, setting loss picture is determined, the setting loss picture is shown with damage component;Determine the setting loss
The difference of picture and template picture, the template picture is shown with the normal component with the damage parts match;According to described
Difference determines damaged condition.It using the above scheme, can be quasi- by the photo for the position of collision or damage component that user provides
It really determines the difference of setting loss picture and template picture, to accurately determine damaged condition, improves convenience.
It is understandable to enable above-mentioned purpose of the invention, feature and beneficial effect to become apparent, with reference to the accompanying drawing to this
The specific embodiment of invention is described in detail.
Referring to Fig.1, Fig. 1 is a kind of flow chart of car damage identification method, the car damage identification method in the embodiment of the present invention
May include
Step S11: setting loss picture is determined, the setting loss picture is shown with damage component;
Step S12: determining the difference of the setting loss picture and template picture, the template picture shown with the damage
The normal component of parts match;
Step S13: damaged condition is determined according to the difference.
In the specific implementation of step S11, setting loss picture, the letter of the upload can be determined by the information that user uploads
It include the picture and information for damaging component in breath;It can also be determined by the picture that scene shoots damage component
Setting loss picture.The setting loss picture only shows the content of damage component as far as possible, thus in subsequent determining setting loss picture
When with the difference of template picture, it is avoided as much as the content interference of non-damaging component.
Specifically, referring to Fig. 2 shows the flow chart for how determining setting loss picture, the step of determining setting loss picture, can
To include step S21 to step S22, each step is described in detail below.
In the step s 21, the damage information that user uploads is obtained, the damage information includes showing the damage component
Uploading pictures.
In specific implementation, when some component of vehicle collides or the damage of other forms, user can use hand
The intelligent terminals such as machine, tablet computer, wearable device shoot damage component, can also use imaging sensor appropriate
(such as camera, camera) shoots damage component, and then image is uploaded.More specifically, service can be uploaded to
Device, cloud platform, car networking server, car networking server etc..
Wherein, the cloud platform (Cloud Platforms) is also known as cloud computing platform, in embodiments of the present invention, cloud
The intelligent terminal that platform can be bound by user carries out information collection, and then is stored, calculated to collected information.
Further, the damage information may include showing the uploading pictures of the damage component, can also include
Position of the damage component on vehicle.It cannot accurately differentiate which position the damage component is when only passing through uploading pictures
When, it can be confirmed according to the position, so as to more accurately determine the damage component.Specifically, such as upper
The place of local damage is only shown in blit piece, can not be located at the tailstock or car door based on its identification damage component, then be needed
It is determined according to the position in user's upload information.
Further, the damage information can also include the information such as vehicle, vehicle brand, be conducive to according to the vehicle
Actual features, accurate judgement setting loss result.
In step S22, the setting loss picture is extracted from the uploading pictures.
In specific implementation, the embodiment of the invention provides a kind of extracts the setting loss picture from the uploading pictures
Method, flow chart are as shown in Figure 3.The method for extracting the setting loss picture may include step S31 to step S32:
Step S31: the damage component is identified from the uploading pictures;
Step S32: according to the damage component identified, the setting loss picture is extracted from the uploading pictures.
In the specific implementation of step S31, institute can be identified from the uploading pictures using the method for manual identified
Damage component is stated, the damage component can also be identified from the uploading pictures using image recognition algorithm.
Wherein, described image recognizer is for handling image, being analyzed and being understood, to identify in various differences
Damage component under damaged condition, such as may include characteristic point image recognition algorithm, the image recognition calculation based on outline identification
Method etc..It should be pointed out that in embodiments of the present invention, specific for image recognition algorithm is selected with no restriction.
In the specific implementation of step S32, can be extracted from the uploading pictures using the method manually extracted described in
Setting loss picture can also extract the setting loss picture using image segmentation algorithm from the uploading pictures.
Wherein, described image partitioning algorithm is used to divide the image into several regions specific, with unique properties, and
The content of damage component is extracted to obtain the setting loss picture, such as may include that the image segmentation based on edge detection is calculated
Method, the image segmentation algorithm based on threshold value, image segmentation algorithm based on region growing etc..
Further, the image segmentation algorithm based on edge detection is a kind of more basic image segmentation algorithm, the calculation
Method passes through the edge feature information that edge detection extracts image first, so based on the edge divide the image into one or
Multiple regions.Wherein the edge of image is used to indicate the place that significant change occurs for gray value in the image, represents in image
The discontinuity of image intensity.
Wherein, more commonly extract picture edge characteristic information algorithm include: Roberts edge detection algorithm,
Sobel edge detection algorithm, Lapalace edge detection algorithm, Prewitt edge detection algorithm, LOG edge detection algorithm,
Canny edge detection algorithm, edge detection algorithm based on wavelet transformation etc..
Preferably due to which wavelet transformation has the characteristic of good locality and multiresolution, based on wavelet transformation
Edge detection algorithm is also more effectively and accurate compared with other several algorithms, and the marginal information extracted is more abundant.
Further, the image segmentation algorithm based on threshold value is another image segmentation algorithm being widely used, should
Algorithm calculates one or more gray threshold by the gray feature of image, and by each pixel of the threshold value and image
Point is compared, and then carries out image segmentation according to the result of comparison.The gray value that the algorithm is suitable for background area is more equal
Dust traces near even situation, such as removal vehicle body scratch.
Image segmentation algorithm based on region growing, for some pixel point sets that will there is certain similar quality in image
Certain region is constituted altogether, and then the region is extracted from the background of picture.The algorithm is suitable for background area
There is the background decorative pattern near the case where larger pixel difference, such as removal vehicle body scratch with the image-region of damage component.
With continued reference to Fig. 1, in the specific implementation of step S12, determine template picture, the template picture shown with institute
The normal component of damage parts match is stated, and then determines the difference of the setting loss picture and template picture.
Specifically, the template picture can be the picture for the normal component being arranged in advance, only show as far as possible
The content of the normal component, to be avoided as much as member perimeter when determining the difference of setting loss picture and template picture
Content interference.
It is possible to further determine the difference of the setting loss picture and template picture using Hash (hash) algorithm, may be used also
To determine the difference of the setting loss picture and template picture otherwise, such as pass through artificial judgment mode.
The embodiment of the invention provides a kind of differences that the setting loss picture and template picture are determined using hash algorithm
Method, as shown in figure 4, each step is described in detail below the method may include step S41 to step S42.
In step S41, cryptographic Hash, the digit of the cryptographic Hash and the template picture are constructed for the setting loss picture
The digit of cryptographic Hash is identical.
In specific implementation, cryptographic Hash can be constructed to template picture in advance, to improve efficiency, it is to be understood that
Cryptographic Hash digit for setting loss picture construction should be identical as the digit of the cryptographic Hash of template picture, thus by step-by-step computation,
Effectively determine the difference of setting loss picture and template picture.
It is the flow chart of the setting loss picture construction cryptographic Hash referring to shown in Fig. 5, may include step S51 to step
S56 is below described in detail each step.
In step s 51, the setting loss picture is converted to the gray scale picture of pre-set level, and is divided into preset number
Cell, there are multiple pixels, each pixel has gray value in each cell.
In specific implementation, by the way that setting loss picture is converted to gray scale picture from color image, it can simplify color, thus
Reduce the complexity that operation is carried out to the difference of setting loss picture and template picture.The gray scale is also known as color range, grayscale or centre
Tone (Half-tone), is used to indicate the bright-dark degree of brightness, and main includes 16 grades, 32 grades, 64 grades.The grey level is got over
The color of height, the gray scale picture is abundanter.
Wherein, as a non-limiting example, it is 64 grades that pre-set level, which can be set, all pixels point in setting loss picture
A total of 64 kinds of colors, i.e., the gray value of each pixel are the numerical value between 0 to 63.
It will be appreciated by those skilled in the art that without departing from the spirit and scope of the present invention, coloured silk can also be directlyed adopt
Chromatic graph piece is analyzed, and is not limited to be converted to gray scale picture.
By the way that setting loss picture is divided into present count destination, it can be based on each cell, in smaller area model
Enclose it is interior setting loss picture is compared with template picture, and then the comparison result of each cell is integrated, completely to be compared
Compared with as a result, compared to directly setting loss picture is compared with template picture, as a result more accurately.
It is understood that the number of cell is more, comparison result is more accurate, but the calculating duration for needing to expend
It is longer.Further, since the digit of cryptographic Hash has the number of the cell of division to determine, the cell that setting loss picture is divided
Number should be identical as the cell number divided to template picture, to effectively determine setting loss picture by step-by-step computation
With the difference of template picture.
As a non-limiting example, it is 64 cells that preset number, which can be set, to obtain 64 cryptographic Hash.
In step S52, in each cell, calculate the average value of the gray value of the multiple pixel, using as
The average gray of each cell.
In step S53, based on the average gray of each cell, the present count destination is calculated
Total average gray.
In specific implementation, the average gray and all 64 cells of each cell can be obtained by calculating
Total average gray.
In step S54, each cell of the gray scale picture is traversed, if the average gray of the cell is big
In or equal to total average gray, the record result of the cell is the first numerical value, otherwise for different from described first
The second value of numerical value.
In specific implementation, the average gray of each cell can be compared with total average gray.If
The average gray of one cell is greater than or equal to total average gray, then is denoted as the first numerical value, flat if it is less than total gray scale
Mean value is then denoted as second value.To which institute is deeper than or is shallower than with the color that the first numerical value and second value embody each cell
State the average color of gray scale picture.
As a unrestricted example, it is 1 that first numerical value, which can be set, and the second value is 0, Huo Zheke
First numerical value is arranged as 0, the second value is 1.
In step S55, using the record result of the present count destination as cryptographic Hash, the position of the cryptographic Hash
Number is identical as the preset number.
In specific implementation, the record result of 64 cells obtained in step S54 can be combined, with structure
At the cryptographic Hash of 64 integers namely the gray scale picture, the also known as fingerprint of the gray scale picture.
Further, the setting loss picture is being switched to using step S51 the gray scale picture of pre-set level, and be divided into
Further include step S56 before present count destination: reducing the size of the setting loss picture.
Specifically, by reducing picture, it is possible to reduce the pixel quantity in picture, to improve operation efficiency.
It is understood that aspect ratio is amplified, reduced or changed to picture, the cryptographic Hash can't be changed,
Further, brightness, contrast are increased or decreased or changes color, will not make the cryptographic Hash that biggish change occur.
With continued reference to Fig. 4, in step S42, the Hash of cryptographic Hash and the template picture based on the setting loss picture
Value, determines the difference.
Specifically, it can be recorded by setting loss picture described in successive appraximation and the record of the template picture as a result, calculating
As a result the number of different cells, using as the difference.
More specifically, can judge to have in 64 of every picture more by comparison setting loss picture and the template picture
The numerical value of few position is different, and using the digit of the different numerical value as the difference.Above-mentioned deterministic process is theoretically equivalent
In calculating " Hamming distance ".
It is understood that the difference is bigger, the difference between two pictures is bigger, and the difference is smaller, two figures
Difference between piece is smaller.
With continued reference to Fig. 1, in the specific implementation of step S13, determine that damaged condition may include root according to the difference
According to the comparison result of the difference and one or more preset thresholds, judge to damage grade.
Fig. 6 is a kind of comparison result according to difference and one or more preset thresholds in the embodiment of the present invention, judges to damage
The flow chart of the step of bad grade may include step S61 to step S65:
Step S61: if the difference judges the damage grade for no damage less than the first preset threshold;
Step S62: if the difference is more than or equal to the first preset threshold and less than the second preset threshold, described in judgement
Damaging grade is slight damage;
Step S63: if the difference is more than or equal to the second preset threshold and is less than third predetermined threshold value, described in judgement
Grade is damaged as moderate damage;
Step S64: if the difference is more than or equal to third predetermined threshold value and less than the 4th preset threshold, described in judgement
Grade is damaged as severe damage;
Step S65: if the difference is more than or equal to the 4th preset threshold, it is judged as null result.
Wherein, first preset threshold is less than second preset threshold, and second preset threshold is less than described the
Three preset thresholds, the third predetermined threshold value are less than the 4th preset threshold.
As a unrestricted example, it is 5 that first preset threshold, which can be set, and second preset threshold is
15, the third predetermined threshold value is 30, and the 4th preset threshold is 40.
Specifically, indicating setting loss picture and template when being only less than 5 numerical value differences in 64 of two pictures
Picture is closely similar, then may determine that the damage grade for no damage;When there are 5 to 15 in 64 of two pictures
When numerical value difference, it is a little different to indicate that setting loss picture exists from template picture, then may determine that the damage grade is slight damage
It is bad;When there are when 15 to 30 numerical value differences, indicate that it is many that setting loss picture and template picture exist in 64 of two pictures
It is mostly different, then it may determine that the damage grade for moderate damage;When there are 30 to 45 numbers in 64 of two pictures
When being worth different, it is a large amount of different to indicate that setting loss picture exists from template picture, then may determine that the damage grade for severe damage;
When there are when 45 or more numerical value differences, indicate setting loss picture with template picture almost not in 64 of two pictures
Together, it to infer that the possibility shown in two pictures is not same position, and then may determine that as null result.
It in embodiments of the present invention, can be by the photo for the position of collision or damage component that user provides, accurately
It determines the difference of setting loss picture and template picture, to accurately determine damaged condition, improves convenience.
Further, the car damage identification method further include: for the damage component, query suggestion buy information and/
Or suggest repair message;It issues the user with the suggestion purchase information and/or suggests repair message.
In existing damage component processing method, the selection of user is usually the tradition shop 4S, other entity maintenace points, electricity
Quotient, O2O platform etc. are easy to be taken advantage of on price, fake products, the maintenance frequency and repair location since information is opaque
It deceives, such as is proposed the new component of replacement when damaging slight, or replace to smaller or informal from producer's network of maintenance and repair of an established trade mark
Network of maintenance and repair.
In embodiments of the present invention, can by cloud platform server to the parameter of a variety of devices in car networking, vehicle and
Data, the parameter of regular brand components and data and electric business platform carry out information integration, to more accurately and reliably inquire
And suggestion purchase information is provided and/or suggests repair message.
The component to need repairing can be provided and how buy replacing for the component using the scheme of the embodiment of the present invention
It changes part and/or how to repair the advisory information of the component original part, to enhance the convenience and economy that vehicle uses.
The embodiment of the present invention also provides a kind of car damage identification device, and structural schematic diagram is referring to Fig. 7.The car damage identification dress
Set may include picture determining module 71, difference determining module 72, damage determining module 73, enquiry module 74 and sending module
75。
Wherein, the picture determining module 71 is adapted to determine that setting loss picture, and the setting loss picture is shown with damage component;
The difference determining module 72 is adapted to determine that the difference of the setting loss picture and template picture, the template picture
Shown with the normal component with the damage parts match;
The damage determining module 73, suitable for determining damaged condition according to the difference.
The enquiry module 74 is suitable for for the damage component, and query suggestion buys information and/or suggests maintenance letter
Breath;
The sending module 75, suitable for issuing the user with the suggestion purchase information and/or suggesting repair message.
A kind of structural schematic diagram of specific embodiment of picture determining module 71 referring to shown in Fig. 8, the picture are true
Cover half block 71 may include acquisition of information submodule 81 and extracting sub-module 82.
Wherein, the acquisition of information submodule 81, the damage information uploaded suitable for obtaining user, the damage information include
Show the uploading pictures of the damage component;
The extracting sub-module 82, suitable for extracting the setting loss picture from the uploading pictures.
Further, the damage information further includes position of the damage component on vehicle.
A kind of structural schematic diagram of specific embodiment of the extracting sub-module 82 referring to shown in Fig. 9, the extraction
Submodule 82 may include identification submodule 91 and picture extracting sub-module 92.
Wherein, the identification submodule 91, suitable for identifying the damage component from the uploading pictures;
The picture extracting sub-module 92, suitable for being mentioned from the uploading pictures according to the damage component identified
Take the setting loss picture.
Further, the identification submodule 91 may include the first identification submodule (not shown), be suitable for using image
Recognizer identifies the damage component from the uploading pictures.
The picture extracting sub-module 92 may include the first picture extracting sub-module (not shown), be suitable for using image point
It cuts algorithm and extracts the setting loss picture from the uploading pictures.
Wherein, described image partitioning algorithm may include: the image segmentation algorithm based on edge detection, the figure based on threshold value
As partitioning algorithm or based on the image segmentation algorithm of region growing.
Further, the difference determining module 72 may include that difference determines submodule (not shown), be suitable for using Kazakhstan
Uncommon algorithm determines the difference of the setting loss picture and template picture.
The difference determines that the structural schematic diagram of submodule is referred to Figure 10, and the difference determines that submodule may include
Cryptographic Hash construction submodule 101 and the first difference determine submodule 102.
Wherein, the cryptographic Hash constructs submodule 101, is suitable for the setting loss picture construction cryptographic Hash, the cryptographic Hash
Digit it is identical as the digit of the cryptographic Hash of the template picture;
First difference determines submodule 102, suitable for cryptographic Hash and the template picture based on the setting loss picture
Cryptographic Hash, determine the difference.
Figure 11 is a kind of structural schematic diagram of specific embodiment of cryptographic Hash construction submodule 101 in Figure 10, the Kazakhstan
Uncommon value construction submodule 101 may include: picture transform subblock 111, mean value calculation submodule 112, overall average calculating
Submodule 113, traversal submodule 114, the first cryptographic Hash construction submodule 115 and diminution submodule 116.
Wherein, the picture transform subblock 111, suitable for the setting loss picture to be converted to the grayscale image of pre-set level
Piece, and it is divided into present count destination, there are multiple pixels, each pixel has gray scale in each cell
Value;
The mean value calculation submodule 112 is suitable in each cell, calculates the gray value of the multiple pixel
Average value, using the average gray as each cell;
The overall average computational submodule 113, suitable for the average gray based on each cell, described in calculating
Total average gray of present count destination;
The traversal submodule 114, suitable for traversing each cell of the gray scale picture, if the ash of the cell
It spends average value and is greater than or equal to total average gray, otherwise it is difference that the record result of the cell, which is the first numerical value,
In the second value of first numerical value;
First cryptographic Hash constructs submodule 115, suitable for using the record result of the present count destination as
Cryptographic Hash, the digit of the cryptographic Hash are identical as the preset number.
The diminution submodule 116, suitable for the setting loss picture is switched to pre-set level in the picture transform subblock
Gray scale picture, and be divided into before present count destination, reduce the size of the setting loss picture.
Further, first difference determines that submodule 102 may include: number computational submodule (not shown), fits
The record of the setting loss picture described in successive appraximation and the template picture is as a result, calculate the number of the different cell of record result
Mesh, using as the difference.
The damage determines that submodule 73 may include grade judging submodule (not shown), be suitable for according to the difference with
The comparison result of one or more preset thresholds judges to damage grade.
Figure 12 is a kind of structural schematic diagram of grade judging submodule, the grade judging submodule in the embodiment of the present invention
It may include: the first estate judging submodule 121, the second grade judging submodule 122, tertiary gradient judging submodule 123,
Four grade judging submodules 124 and the 5th grade judging submodule 125.
Wherein, the first estate judging submodule 121 is suitable for when the difference is less than the first preset threshold, judgement
The damage grade is no damage;
The second grade judging submodule 122 is suitable for being more than or equal to the first preset threshold and less than the when the difference
When two preset thresholds, judge the damage grade for slight damage;
The tertiary gradient judging submodule 123 is suitable for being more than or equal to the second preset threshold and less than the when the difference
When three preset thresholds, judge the damage grade for moderate damage;
The fourth estate judging submodule 124 is suitable for being more than or equal to third predetermined threshold value and less than the when the difference
When four preset thresholds, judge the damage grade for severe damage;
The 5th grade judging submodule 125 is suitable for being judged as when the difference is more than or equal to four preset thresholds
Null result;
Wherein, first preset threshold is less than second preset threshold, and second preset threshold is less than described the
Three preset thresholds, the third predetermined threshold value are less than the 4th preset threshold.
Further, the car damage identification device further includes enquiry module (not shown) and sending module (not shown).
Wherein, the enquiry module is suitable for for the damage component, and query suggestion buys information and/or suggests maintenance
Information;
The sending module, suitable for issuing the user with the suggestion purchase information and/or suggesting repair message.
More detailed contents about the car damage identification device please refer to fixed about vehicle above and shown in Fig. 1 to Fig. 6
The associated description of damage method, details are not described herein again.
The embodiment of the invention also provides a kind of computer readable storage mediums, are stored thereon with computer instruction, described
The step of computer instruction executes above-mentioned car damage identification method when running.The computer readable storage medium can be CD,
Mechanical hard disk, solid state hard disk etc..
The embodiment of the invention also provides a kind of terminal, including memory and processor, energy is stored on the memory
Enough computer instructions run on the processor, the processor executes above-mentioned vehicle when running the computer instruction fixed
The step of damage method.
In specific implementation, the terminal can be vehicle, intelligent terminal, cloud platform, car networking server, Internet of Things clothes
Business device etc..Wherein, the intelligent terminal external can be coupled to vehicle, or be integrated in the vehicle, for example, vehicle
Car running computer.
Wherein, the cloud platform (Cloud Platforms) is also known as cloud computing platform, in embodiments of the present invention, cloud
The intelligent terminal that platform can be bound by user carries out information collection, and then is stored, calculated to collected information.
Although present disclosure is as above, present invention is not limited to this.Anyone skilled in the art are not departing from this
It in the spirit and scope of invention, can make various changes or modifications, therefore protection scope of the present invention should be with claim institute
Subject to the range of restriction.
Claims (32)
1. a kind of car damage identification method, which comprises the following steps:
Determine setting loss picture, the setting loss picture is shown with damage component;
Determine the difference of the setting loss picture and template picture, the template picture shown with the damage parts match just
Normal component;
Damaged condition is determined according to the difference.
2. car damage identification method according to claim 1, which is characterized in that the determining setting loss picture includes:
The damage information that user uploads is obtained, the damage information includes showing the uploading pictures of the damage component;
The setting loss picture is extracted from the uploading pictures.
3. car damage identification method according to claim 2, which is characterized in that the damage information further includes the damage portion
Position of the part on vehicle.
4. car damage identification method according to claim 2, which is characterized in that extract the setting loss from the uploading pictures
Picture includes:
The damage component is identified from the uploading pictures;
According to the damage component identified, the setting loss picture is extracted from the uploading pictures.
5. car damage identification method according to claim 4, which is characterized in that identify the damage from the uploading pictures
Bad component includes:
The damage component is identified from the uploading pictures using image recognition algorithm.
6. car damage identification method according to claim 4, which is characterized in that extract the setting loss from the uploading pictures
Picture includes:
The setting loss picture is extracted from the uploading pictures using image segmentation algorithm.
7. car damage identification method according to claim 6, which is characterized in that described image partitioning algorithm includes: based on side
Image segmentation algorithm, the image segmentation algorithm based on threshold value or the image segmentation algorithm based on region growing of edge detection.
8. car damage identification method according to claim 1, which is characterized in that determine the setting loss picture and template picture
Difference includes:
The difference of the setting loss picture and template picture is determined using hash algorithm.
9. car damage identification method according to claim 8, which is characterized in that described to determine the setting loss using hash algorithm
The difference of picture and template picture includes:
Cryptographic Hash, the digit phase of the digit of the cryptographic Hash and the cryptographic Hash of the template picture are constructed for the setting loss picture
Together;
The cryptographic Hash of cryptographic Hash and the template picture based on the setting loss picture, determines the difference.
10. car damage identification method according to claim 9, which is characterized in that construct cryptographic Hash packet for the setting loss picture
It includes:
The setting loss picture is converted to the gray scale picture of pre-set level, and is divided into present count destination, Mei Gedan
There are multiple pixels, each pixel has gray value in first lattice;
In each cell, the average value of the gray value of the multiple pixel is calculated, using the gray scale as each cell
Average value;
Based on the average gray of each cell, total average gray of the present count destination is calculated;
Each cell of the gray scale picture is traversed, if the average gray of the cell is greater than or equal to total ash
Average value is spent, otherwise it is the second value different from first numerical value that the record result of the cell, which is the first numerical value,;
Using the record result of the present count destination as cryptographic Hash, the digit of the cryptographic Hash and the preset number
It is identical.
11. car damage identification method according to claim 10, which is characterized in that the setting loss picture is being switched to default grade
Other gray scale picture, and be divided into before present count destination, cryptographic Hash is constructed for the setting loss picture further include:
Reduce the size of the setting loss picture.
12. car damage identification method according to claim 9, which is characterized in that cryptographic Hash based on the setting loss picture with
The cryptographic Hash of the template picture determines that the difference includes:
The record of setting loss picture described in successive appraximation and the template picture is as a result, calculate the number of the different cell of record result
Mesh, using as the difference.
13. car damage identification method according to claim 1, which is characterized in that determine damaged condition packet according to the difference
It includes:
According to the comparison result of the difference and one or more preset thresholds, judge to damage grade.
14. car damage identification method according to claim 13, which is characterized in that pre- according to the difference and one or more
If the comparison result of threshold value, judge that damaging grade includes:
If the difference judges the damage grade for no damage less than the first preset threshold;
If the difference is more than or equal to the first preset threshold and less than the second preset threshold, judge that the damage grade is light
Degree damage;
If the difference is more than or equal to the second preset threshold and is less than third predetermined threshold value, judge that the damage grade is
Degree damage;
If the difference is more than or equal to third predetermined threshold value and less than the 4th preset threshold, judge that the damage grade is attached most importance to
Degree damage;
If the difference is more than or equal to the 4th preset threshold, it is judged as null result;
Wherein, first preset threshold is less than second preset threshold, and it is pre- that second preset threshold is less than the third
If threshold value, the third predetermined threshold value is less than the 4th preset threshold.
15. car damage identification method according to claim 1, which is characterized in that further include:
For the damage component, query suggestion buys information and/or suggests repair message;
It issues the user with the suggestion purchase information and/or suggests repair message.
16. a kind of car damage identification device characterized by comprising
Picture determining module is adapted to determine that setting loss picture, and the setting loss picture is shown with damage component;
Difference determining module, is adapted to determine that the difference of the setting loss picture and template picture, the template picture shown with institute
State the normal component of damage parts match;
Determining module is damaged, suitable for determining damaged condition according to the difference.
17. car damage identification device according to claim 16, which is characterized in that the picture determining module includes:
Acquisition of information submodule, the damage information uploaded suitable for obtaining user, the damage information include showing the damage
The uploading pictures of component;
Extracting sub-module, suitable for extracting the setting loss picture from the uploading pictures.
18. car damage identification device according to claim 17, which is characterized in that the damage information further includes the damage
Position of the component on vehicle.
19. car damage identification device according to claim 17, which is characterized in that the extracting sub-module includes:
Submodule is identified, suitable for identifying the damage component from the uploading pictures;
Picture extracting sub-module, suitable for extracting the setting loss from the uploading pictures according to the damage component identified
Picture.
20. car damage identification device according to claim 19, which is characterized in that the identification submodule includes:
First identification submodule, suitable for identifying the damage component from the uploading pictures using image recognition algorithm.
21. car damage identification device according to claim 19, which is characterized in that the picture extracting sub-module includes:
First picture extracting sub-module, suitable for extracting the setting loss picture from the uploading pictures using image segmentation algorithm.
22. car damage identification device according to claim 21, which is characterized in that described image partitioning algorithm includes: to be based on
The image segmentation algorithm of edge detection, the image segmentation algorithm based on threshold value or the image segmentation algorithm based on region growing.
23. car damage identification device according to claim 16, which is characterized in that the difference determining module includes:
Difference determines submodule, suitable for determining the difference of the setting loss picture and template picture using hash algorithm.
24. car damage identification device according to claim 23, which is characterized in that the difference determines that submodule includes:
Cryptographic Hash constructs submodule, is suitable for the setting loss picture construction cryptographic Hash, the digit of the cryptographic Hash and the template
The digit of the cryptographic Hash of picture is identical;
First difference determines submodule, suitable for the cryptographic Hash of cryptographic Hash and the template picture based on the setting loss picture, really
The fixed difference.
25. car damage identification device according to claim 24, which is characterized in that the cryptographic Hash constructs submodule and includes:
Picture transform subblock suitable for the setting loss picture to be converted to the gray scale picture of pre-set level, and is divided into default
Number destination, each cell is interior to have multiple pixels, and each pixel has gray value;
Mean value calculation submodule is suitable in each cell, calculates the average value of the gray value of the multiple pixel, with
Average gray as each cell;
Overall average computational submodule calculates the preset number suitable for the average gray based on each cell
Total average gray of cell;
Submodule is traversed, suitable for traversing each cell of the gray scale picture, if the average gray of the cell is big
In or equal to total average gray, the record result of the cell is the first numerical value, otherwise for different from described first
The second value of numerical value;
First cryptographic Hash constructs submodule, described suitable for using the record result of the present count destination as cryptographic Hash
The digit of cryptographic Hash is identical as the preset number.
26. car damage identification device according to claim 25, which is characterized in that the cryptographic Hash construction submodule also wraps
It includes:
Submodule is reduced, suitable for the setting loss picture to be switched to the gray scale picture of pre-set level in the picture transform subblock,
And it is divided into before present count destination, reduces the size of the setting loss picture.
27. car damage identification device according to claim 24, which is characterized in that first difference determines submodule packet
It includes:
Number computational submodule is recorded suitable for setting loss picture described in successive appraximation and the record of the template picture as a result, calculating
As a result the number of different cells, using as the difference.
28. car damage identification device according to claim 16, which is characterized in that the damage determines that submodule includes:
Grade judging submodule judges damage etc. suitable for the comparison result according to the difference and one or more preset thresholds
Grade.
29. car damage identification device according to claim 28, which is characterized in that the grade judging submodule includes:
The first estate judging submodule is suitable for when the difference is less than the first preset threshold, judges the damage grade for nothing
Damage;
Second grade judging submodule is suitable for being more than or equal to the first preset threshold and less than the second preset threshold when the difference
When, judge the damage grade for slight damage;
Tertiary gradient judging submodule is suitable for when the difference is more than or equal to the second preset threshold and is less than third predetermined threshold value
When, judge the damage grade for moderate damage;
Fourth estate judging submodule is suitable for being more than or equal to third predetermined threshold value and less than the 4th preset threshold when the difference
When, judge the damage grade for severe damage;
5th grade judging submodule is suitable for being judged as null result when the difference is more than or equal to four preset thresholds;
Wherein, first preset threshold is less than second preset threshold, and it is pre- that second preset threshold is less than the third
If threshold value, the third predetermined threshold value is less than the 4th preset threshold.
30. car damage identification device according to claim 16, which is characterized in that further include:
Enquiry module is suitable for for the damage component, and query suggestion buys information and/or suggests repair message;
Module is issued, suitable for issuing the user with the suggestion purchase information and/or suggesting repair message.
31. a kind of computer readable storage medium, is stored thereon with computer instruction, which is characterized in that the computer instruction
Perform claim requires the step of any one of 1 to 15 car damage identification method when operation.
32. a kind of terminal, including memory and processor, be stored on the memory to run on the processor
Computer instruction, which is characterized in that perform claim requires any one of 1 to 15 institute when the processor runs the computer instruction
The step of stating car damage identification method.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710443445.0A CN109146834A (en) | 2017-06-13 | 2017-06-13 | Car damage identification method and device, computer readable storage medium, terminal |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710443445.0A CN109146834A (en) | 2017-06-13 | 2017-06-13 | Car damage identification method and device, computer readable storage medium, terminal |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109146834A true CN109146834A (en) | 2019-01-04 |
Family
ID=64829933
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710443445.0A Pending CN109146834A (en) | 2017-06-13 | 2017-06-13 | Car damage identification method and device, computer readable storage medium, terminal |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109146834A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2020206900A1 (en) * | 2019-04-09 | 2020-10-15 | 平安科技(深圳)有限公司 | Vehicle loss assessment method and apparatus, computer device and storage medium |
CN112085612A (en) * | 2020-11-13 | 2020-12-15 | 深圳壹账通智能科技有限公司 | Vehicle total loss detection method and device, terminal equipment and storage medium |
CN116434047A (en) * | 2023-03-29 | 2023-07-14 | 邦邦汽车销售服务(北京)有限公司 | Vehicle damage range determining method and system based on data processing |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103493051A (en) * | 2011-02-25 | 2014-01-01 | 奥达特克斯有限责任公司 | System and method for estimating collision damage to a car |
CN104679788A (en) * | 2013-12-02 | 2015-06-03 | 中国移动通信集团广东有限公司 | Image processing method and device as well as terminal equipment |
CN105488789A (en) * | 2015-11-24 | 2016-04-13 | 大连楼兰科技股份有限公司 | Grading damage assessment method for automobile part |
CN105912643A (en) * | 2016-04-08 | 2016-08-31 | 浙江理工大学 | Image retrieval method based on content improved Average Hash |
CN106296118A (en) * | 2016-08-03 | 2017-01-04 | 深圳市永兴元科技有限公司 | Car damage identification method based on image recognition and device |
CN106600612A (en) * | 2016-12-27 | 2017-04-26 | 重庆大学 | Damage identification and detection method for electric automobile before and after renting |
-
2017
- 2017-06-13 CN CN201710443445.0A patent/CN109146834A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103493051A (en) * | 2011-02-25 | 2014-01-01 | 奥达特克斯有限责任公司 | System and method for estimating collision damage to a car |
CN104679788A (en) * | 2013-12-02 | 2015-06-03 | 中国移动通信集团广东有限公司 | Image processing method and device as well as terminal equipment |
CN105488789A (en) * | 2015-11-24 | 2016-04-13 | 大连楼兰科技股份有限公司 | Grading damage assessment method for automobile part |
CN105912643A (en) * | 2016-04-08 | 2016-08-31 | 浙江理工大学 | Image retrieval method based on content improved Average Hash |
CN106296118A (en) * | 2016-08-03 | 2017-01-04 | 深圳市永兴元科技有限公司 | Car damage identification method based on image recognition and device |
CN106600612A (en) * | 2016-12-27 | 2017-04-26 | 重庆大学 | Damage identification and detection method for electric automobile before and after renting |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2020206900A1 (en) * | 2019-04-09 | 2020-10-15 | 平安科技(深圳)有限公司 | Vehicle loss assessment method and apparatus, computer device and storage medium |
CN112085612A (en) * | 2020-11-13 | 2020-12-15 | 深圳壹账通智能科技有限公司 | Vehicle total loss detection method and device, terminal equipment and storage medium |
CN112085612B (en) * | 2020-11-13 | 2021-03-16 | 深圳壹账通智能科技有限公司 | Vehicle total loss detection method and device, terminal equipment and storage medium |
WO2022100455A1 (en) * | 2020-11-13 | 2022-05-19 | 深圳壹账通智能科技有限公司 | Vehicle total loss detection method and apparatus, terminal device and storage medium |
CN116434047A (en) * | 2023-03-29 | 2023-07-14 | 邦邦汽车销售服务(北京)有限公司 | Vehicle damage range determining method and system based on data processing |
CN116434047B (en) * | 2023-03-29 | 2024-01-09 | 邦邦汽车销售服务(北京)有限公司 | Vehicle damage range determining method and system based on data processing |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109583483B (en) | Target detection method and system based on convolutional neural network | |
CN107203990A (en) | A kind of labeling damage testing method based on template matches and image quality measure | |
CN108846401A (en) | Commodity detect terminal, method, system and computer equipment, readable medium | |
CN106228528B (en) | A kind of multi-focus image fusing method based on decision diagram and rarefaction representation | |
CN103093208B (en) | A kind of method and system of fruits and vegetables identification | |
CN106446952A (en) | Method and apparatus for recognizing score image | |
CN105184225B (en) | A kind of multinational banknote image recognition methods and device | |
CN109146834A (en) | Car damage identification method and device, computer readable storage medium, terminal | |
CN113240623B (en) | Pavement disease detection method and device | |
CN111784675A (en) | Method and device for processing article texture information, storage medium and electronic equipment | |
CN111415339A (en) | Image defect detection method for complex texture industrial product | |
CN109977251A (en) | A method of building identifies commodity based on RGB histogram feature | |
CN115082776A (en) | Electric energy meter automatic detection system and method based on image recognition | |
CN109389165A (en) | Oil level gauge for transformer recognition methods based on crusing robot | |
CN109284759A (en) | One kind being based on the magic square color identification method of support vector machines (svm) | |
CN113780484B (en) | Industrial product defect detection method and device | |
CN109784328A (en) | Position method, terminal and the computer readable storage medium of bar code | |
CN113505629A (en) | Intelligent storage article recognition device based on light weight network | |
JP5931646B2 (en) | Image processing device | |
Alaei et al. | Document image quality assessment based on texture similarity index | |
CN108805838A (en) | A kind of image processing method, mobile terminal and computer readable storage medium | |
WO2024016632A1 (en) | Bright spot location method, bright spot location apparatus, electronic device and storage medium | |
CN116597438A (en) | Improved fruit identification method and system based on Yolov5 | |
CN106446832B (en) | Video-based pedestrian real-time detection method | |
CN207636505U (en) | A kind of shoes mould surface damage detection device |
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 |