CN113392455A - House type graph scale detection method and device based on deep learning and electronic equipment - Google Patents
House type graph scale detection method and device based on deep learning and electronic equipment Download PDFInfo
- Publication number
- CN113392455A CN113392455A CN202110656784.3A CN202110656784A CN113392455A CN 113392455 A CN113392455 A CN 113392455A CN 202110656784 A CN202110656784 A CN 202110656784A CN 113392455 A CN113392455 A CN 113392455A
- Authority
- CN
- China
- Prior art keywords
- scale
- house type
- area
- type graph
- end point
- 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.)
- Granted
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 29
- 238000013135 deep learning Methods 0.000 title claims abstract description 23
- 238000000034 method Methods 0.000 claims abstract description 30
- 238000010276 construction Methods 0.000 claims abstract description 9
- 238000004590 computer program Methods 0.000 claims description 11
- 238000012545 processing Methods 0.000 claims description 10
- 238000006243 chemical reaction Methods 0.000 claims description 9
- 238000012795 verification Methods 0.000 claims description 7
- 238000004364 calculation method Methods 0.000 claims description 6
- 230000002093 peripheral effect Effects 0.000 claims description 4
- 230000007797 corrosion Effects 0.000 claims description 3
- 238000005260 corrosion Methods 0.000 claims description 3
- 230000010339 dilation Effects 0.000 claims description 2
- 230000003628 erosive effect Effects 0.000 claims description 2
- 239000000126 substance Substances 0.000 claims 1
- 238000005034 decoration Methods 0.000 abstract description 6
- 238000013461 design Methods 0.000 abstract description 4
- 238000011161 development Methods 0.000 abstract description 3
- 238000010586 diagram Methods 0.000 description 7
- 238000004891 communication Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 230000003416 augmentation Effects 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/10—Geometric CAD
- G06F30/13—Architectural design, e.g. computer-aided architectural design [CAAD] related to design of buildings, bridges, landscapes, production plants or roads
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Geometry (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Computer Hardware Design (AREA)
- General Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Medical Informatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Artificial Intelligence (AREA)
- Architecture (AREA)
- Civil Engineering (AREA)
- Structural Engineering (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Image Analysis (AREA)
Abstract
The invention provides a method and a device for detecting a house type graph scale based on deep learning and electronic equipment. The method comprises the steps of obtaining the position of a ruler area of a user pattern; acquiring a physical size area position in the scale area; splitting and identifying physical dimension values within the physical dimension region; constructing a scale end point template pair to be matched with the scale area, acquiring the position of the scale end point template pair when the matching degree exceeds a preset threshold value, and determining the position of the scale end point pair; determining the pixel number of the scale according to the position of the scale end point pair; pairing the physical size and the scale and determining the house type graph scale according to the number of pixels of the scale; and verifying the scale accuracy of the house type graph. The invention can improve the precision rate of the detection of the scale, thereby improving the design efficiency of a home decoration designer, opening a full-automatic link of house type identification and accelerating the development of digital construction of the home decoration industry.
Description
Technical Field
The invention relates to the field of home decoration design, in particular to a method and a device for detecting a house type graph scale based on deep learning and electronic equipment.
Background
At present, automatic identification of house type pictures of house type businesses, automatic layout of house and digital construction of house type all need to detect the physical dimensions of all components in the house type, and the detection of the scale of the house type pictures is particularly important. In the prior art, the scale of the house type graph is obtained by calculation mainly through manual input, the conventional OCR technology is adopted, the former has the problems of low efficiency, low accuracy, incapability of completing a full-automatic link of house type identification and the like in a manual mode, and the latter has the problems of insufficient generalization capability of scale area positioning, low scale identification rate and the like.
Disclosure of Invention
In view of the above drawbacks of the prior art, an object of the present invention is to provide a method, an apparatus and an electronic device for detecting a house type graph scale based on deep learning, which are used to solve the above problems in the prior art.
In order to achieve the above and other related objects, the present invention provides a method, an apparatus and an electronic device for detecting a house type graph scale based on deep learning, wherein the method comprises: acquiring the position of a scale region of a house-type map; acquiring a physical size area position in the scale area; identifying a physical dimension value within the physical dimension region; constructing a scale end point template pair, matching the scale end point template pair with the scale area, acquiring the position of the scale end point template pair when the matching degree exceeds a preset threshold value, and calculating the position of the scale end point pair; calculating the pixel number of the scale according to the position of the scale end point pair; matching the physical size and the scale and calculating the house type graph scale according to the number of pixels of the scale; and verifying the scale accuracy of the house type graph.
In an embodiment of the present invention, the method further includes: converting the floor-type map into a gray-scale map and calculating a pixel histogram of the gray-scale map; determining the gray-scale image pixel conversion threshold value according to the pixel histogram; and converting the gray-scale image into a black-white image according to the pixel conversion threshold value.
In an embodiment of the present invention, the method further includes detecting a rectangular region where the house type diagram is located according to the trained house type diagram detection model and obtaining vertex coordinates of the region; expanding the rectangular area where the house type graph is located to the peripheral area according to the vertex coordinates; performing expansion corrosion treatment on the expansion area, wherein the expansion area after treatment is the scale area; and acquiring the coordinate position of the scale area.
In an embodiment of the present invention, the method further includes: detecting the physical size in the scale area and positioning to obtain the position coordinate of the physical size; resolving numbers in the physical dimensions by erosion dilation; and predicting the numerical corresponding value in the physical dimension according to a pre-established numerical prediction model.
In an embodiment of the present invention, the method further includes: constructing a corresponding ruler endpoint template pair according to the type of the end point of the house type icon ruler; traversing the scale region of the house-type figure through the scale endpoint template, and calculating the matching degree of the scale endpoint template and the scale region according to the number of the overlapped pixel values of the scale endpoint template and the scale region of the house-type figure.
In an embodiment of the present invention, the method further includes: matching the physical size value with the scale according to the position of the connecting line of the physical size area position and the scale end point pair; determining a scale of a corresponding scale according to the physical size and the scale pixel number which are successfully matched; and determining the scale of the house type graph according to the scale value with the deviation value smaller than a preset threshold value, wherein the preset threshold value can be dynamically adjusted.
In an embodiment of the present invention, the method further includes: comparing and determining the product of the house type graph scale and the scale pixel number with the deviation value of the scale physical dimension value; and when the deviation value of the scale is smaller than the preset deviation value threshold, the scale accuracy of the house type graph meets the standard, otherwise, the scale is detected again.
To achieve the above and other related objects, the present invention provides a device for detecting a house type scale based on deep learning, the device comprising: the position acquisition module is used for positioning and acquiring the position of the scale area, the position of the physical size area and the position of the scale end point pair; the text recognition module is used for processing and predicting the physical size value in the physical size area; the construction module is used for constructing the scale endpoint template pair according to the scale endpoint types; the matching module is used for matching the scale endpoint template pair with the scale area; the calculation module is used for calculating the scale of the house type graph; and the verification module is used for verifying the accuracy of the scale of the house type graph.
To achieve the above and other related objects, the present invention provides a computer-readable storage medium, in which a computer program is stored, and when the computer program is loaded and executed by a processor, the method for detecting a scale of a user-type graph based on deep learning is implemented.
To achieve the above and other related objects, the present invention provides an electronic device, comprising: a processor, a memory; wherein the memory is for storing a computer program; the processor is used for loading and executing the computer program to enable the electronic equipment to execute the deep learning-based house type graph scale detection method.
As described above, according to the house type graph scale detection method, device and electronic device based on deep learning provided by the invention, the scale region is extracted through target detection, the precision rate and the calculation speed of scale detection are improved, the recall rate of detection is greatly improved through text recognition and pixel level template matching positioning, and the accuracy rate of scale detection is improved through self-verification of detection results by adopting reverse statistics, so that the design efficiency of a house ornamentation designer is improved, a full-automatic link of house type recognition is opened, and the development of digital construction of the house ornamentation industry is accelerated.
Drawings
Fig. 1 is a flowchart illustrating a method for detecting a house type graph scale based on deep learning according to an embodiment of the present invention.
Fig. 2 is a schematic diagram of a house type graph to be detected based on the deep learning house type graph scale detection method in an embodiment of the present invention.
Fig. 3 is a schematic structural diagram of a device for detecting a house type scale based on deep learning according to an embodiment of the present invention.
Fig. 4 is a schematic diagram of an electronic device according to an embodiment of the invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. The components of embodiments of the present invention generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations.
Thus, the following detailed description of the embodiments of the present invention, presented in the figures, is not intended to limit the scope of the invention, as claimed, but is merely representative of selected embodiments of the invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
In the description of the present invention, it should be noted that the terms "first", "second", and the like are used for distinguishing between the descriptions and are not to be construed as indicating or implying relative importance.
The current house type graph generally refers to a plane space layout graph, and the size marking of a space peripheral wall body is provided around the house type graph generally, so that the length information of the whole house of a user is facilitated. However, in the automatic home layout and digital home decoration construction, the physical dimensions of all components in the house type are required to be calculated, and the scale of the house type graph is required to be detected for calculating the physical dimensions of all the components.
Referring to fig. 1, the present embodiment provides a method for detecting a house type graph scale based on deep learning, including the following steps:
s11: the position of the scale region of the house figure is detected.
Specifically, the house type image is subjected to binarization, wherein binarization is to convert the color house type image into a gray scale image, and then the gray scale image is further converted into a black and white image according to a pixel conversion threshold value. It should be noted that, due to the diversity of the floor plan, the pixel conversion threshold needs to be adaptive to all floor plan. Specifically, the pixel histogram after the gray processing of the user-type image is calculated first, then the distribution condition of the pixel histogram is counted, and finally the pixel value corresponding to the previous pit of the white pixel 255 is found, and the value is used as the pixel conversion threshold, that is, the pixel value greater than the conversion threshold is converted into the white pixel value 255, and the pixel value less than the conversion threshold is converted into the black pixel value 0, so that the gray image is converted into the black-and-white image.
Further, the rectangular house type region detection is carried out on the house type graph through the trained Efficientdet target detection model, namely, the rectangular house type region is extracted according to the upper left corner coordinate and the lower right corner coordinate of the house type region of the house type graph, and the rectangular house type region is respectively expanded upwards and downwards or expanded leftwards and rightwards according to the upper, lower, left and right sides of the rectangular frame of the detected house type region. Preferably, two regions are formed by expanding upwards and rightwards to the upper edge and the lower edge of the house type graph, then closed operation is realized in the regions through opencv, namely expansion is performed firstly and then corrosion is performed, gray and black pixels are expanded and spread outwards firstly and then contract inwards integrally, a plurality of sections of scales are formed into an upper whole scale region and a lower whole scale region, as shown in fig. 2, rectangular regions corresponding to an upper region H1 and a lower region H2 of the house type graph are the house type icon scale regions. And finally, acquiring coordinates of the upper left corner and the lower right corner of the scale area.
S12: acquiring a physical size area position in the scale area.
Specifically, a digital target detection model is established through a deep learning method and digital sample augmentation according to numbers of different fonts existing in different house type graphs, target detection is carried out on physical dimensions in a ruler area, and position coordinates of the physical dimensions are obtained through positioning, and preferably coordinates of the upper left corner and the lower right corner of the physical dimension area are obtained. As shown in fig. 2, the area corresponding to the house type diagram B is an area where a physical size in the house type diagram is located.
S13: identifying a physical dimension value within the physical dimension region.
Specifically, the number in the region is subjected to splitting recognition after the physical size region is acquired according to S12. Specifically, all the numbers in the located physical size area are split through open operation of an opencv library, namely, firstly corroding and then expanding, inwards contracting and then expanding gray and black pixels.
Furthermore, semantic features of the region are extracted from each number through a convolutional neural network, a number prediction model is built to predict the probability that the number belongs to 0-9, the probability that the number belongs to 0-9 is compared, and the number with the highest probability is selected as the predicted value of the number. The probability that the number belongs to 0-9 is predicted, for example, by the softmax function.
S14: and constructing a scale end point template pair, matching the scale end point template pair with the scale area, acquiring the position of the scale end point template pair when the matching degree exceeds a preset threshold value, and calculating the position of the scale end point pair.
Specifically, the end points of the house type icon ruler are classified according to the house type pattern of the current online platform, such as the ruler end types of the K type, the I type and the like. And counting the pixel quantity occupied by the ruler endpoint area in the house type picture, and constructing different types of ruler endpoint template pairs according to the pixel quantity. The scale end point template pair consists of a pair of opposite scale end point templates, namely the left end and the right end of the scale, and a pair of scale end points form a section of scale, as shown by the scale end points corresponding to A1 and A2 in FIG. 2.
Further, according to step S11, the layout has been binarized, so that a pair of 2-dimensional arrays is used to construct a ruler end point template pair, where a blank area is represented by a pixel value 255 and a ruler is represented by a pixel value 0. For example, according to big data statistics, the scale endpoint region in the house type graph occupies no more than 10x 10 of pixels, so a 10x 10 two-dimensional array can be constructed as a scale endpoint template pair.
Further, each group of template pairs is used for traversing the scale area of the binarized house type graph, the number of the pixel value of 0 in the template array and the repeated number of the pixel value of 0 in the scale area are calculated in the traversing process, and the matching degree is calculated through the repeated number of the pixel value of 0 and the total number of the template pixels.
Further, when the matching degree is greater than a preset threshold value, the position coordinates of the scale end point template pair are obtained, and the position of the corresponding scale end point pair is calculated according to the position. Wherein the preset threshold is obtained through a large number of tests.
S15: and determining the pixel number of the scale according to the positions of the scale end point pairs.
Specifically, the number of pixel points corresponding to the segment of the scale is calculated through the position coordinates of the scale endpoint pairs, if the household-type icon scale region is detected by expanding the upper region and the lower region of the household-type graph, the difference value of two horizontal coordinates of the scale endpoint pairs is the number of scale pixels, and if the household-type icon scale region is detected by expanding the left region and the right region of the household-type graph, the difference value of two vertical coordinates of the scale endpoint pairs is the number of scale pixels.
S16: and pairing the physical size and the scale and determining the house type graph scale according to the number of pixels of the scale.
Specifically, the physical size value and the scale are paired according to the physical size region position acquired in step S12 and the scale end point pair position acquired in step S15.
Further, if the physical dimension area rectangular frame and the scale end point pair are located in the same scale area and the physical dimension area rectangular frame is located between the scale end point pair, the corresponding scale and the physical dimension value are successfully paired. That is, if the section of the scale is located in the scale area or the lower scale area on the house type figure, the abscissa of the rectangular frame of the physical size area needs to be between the end point pair abscissas of the scale, and the corresponding scale and the physical size value are successfully matched; if the scale is in the left scale area or the right scale area of the house-type picture, the vertical coordinate of the rectangular frame of the physical size area is required to be between the end point pair vertical coordinate of the scale, and the corresponding scale and the physical size value are successfully matched.
Furthermore, the scales of the corresponding scales are calculated according to the successfully matched physical dimension values and the number of pixels of the scales, and therefore a plurality of scales in the user-type graph can be obtained.
It should be noted that, because the calculation of the number of the scale pixels may have a certain error, a plurality of calculated scales may also have a certain deviation value. And when the deviation value is larger than a threshold value, filtering the scale, and calculating the scale of the house type graph according to the residual scale, wherein the threshold value is comprehensively determined according to the statistical principle.
S17: and verifying the scale accuracy of the house type graph.
Specifically, the accuracy of the house type scale is reversely verified according to the house type scale obtained in step S16 and the number of pixels corresponding to the scale obtained in step S15, that is, the number of pixels corresponding to each segment of scale is compared with the product of the house type scale obtained in step S16 and the physical size value of the corresponding scale, when the deviation between the product of a segment of scale and the corresponding physical size value is smaller than the preset deviation value threshold, the accuracy of the house type scale meets the standard, otherwise, scale detection needs to be performed again. As shown in fig. 2, C represents the scale of the house type graph, the scale is 19.45, the number of pixels between the scales A1a2 is 150 pixels calculated in step S15, the product of the number of pixels of the scale and the scales A1a2 is 2918, the physical size value 2899 of the scale is compared, the deviation value is 19, if the preset deviation value threshold is set to 10 in the scale verification, the deviation value of the scales A1a2 is greater than the preset value, the deviation value of the next scale section is continuously verified, and if all the deviation values of the scales are greater than the preset value, the scale detection is performed again.
All or part of the steps for implementing the above method embodiments may be performed by hardware associated with a computer program. Based upon such an understanding, the present invention also provides a computer program product comprising one or more computer instructions. The computer instructions may be stored in a computer readable storage medium. The computer-readable storage medium can be any available medium that a computer can store or a data storage device, such as a server, a data center, etc., that is integrated with one or more available media.
Referring to fig. 3, the present embodiment provides a device 30 for detecting a house type scale based on deep learning, which is installed in an electronic device as a device to execute the method for detecting a house type scale based on deep learning in the foregoing method embodiment during operation. Since the technical principle of the embodiment of the apparatus is similar to that of the embodiment of the method, repeated description of the same technical details is omitted.
The device 30 for detecting a house type graph scale based on deep learning of the present embodiment specifically includes: a position acquisition module 31, a text recognition module 32, a construction module 33, a matching module 34, a calculation module 35, and a verification module 36. The position obtaining module 31 is configured to locate and obtain the position of the scale region, the position of the physical size region, and the position of the scale end point pair; the text recognition module 32 is used for processing and predicting the physical size value in the physical size area; the construction module 33 is configured to construct the scale endpoint template pair according to the scale endpoint type; a matching module 34 is configured to match the pair of scale endpoint templates to the scale region; the calculating module 35 is used for calculating the scale of the house type graph; the verification module 36 is used for verifying the accuracy of the house type graph scale.
Those skilled in the art should understand that the division of the modules in the embodiment of fig. 3 is only a logical division, and the actual implementation can be fully or partially integrated into one or more physical entities. The modules can be embedded in a hardware form or independent of a processor in the computer device, can also be stored in a memory in the computer device in a software form, can also be realized in a form that part of the modules call software through a processing element, and can be realized in a hardware form. For example, the matching module 34 may be a separate processing element, or may be implemented by being integrated in a chip, or may be stored in a memory in the form of program code, and the function of the matching module 34 is called and executed by a certain processing element. Other modules are implemented similarly. The processing element described herein may be an integrated circuit having signal processing capabilities. In implementation, each step of the above method or each module above may be implemented by an integrated logic circuit of hardware in a processor element or an instruction in the form of software.
Referring to fig. 4, the embodiment provides an electronic device, which may be a portable computer, a smart phone, a tablet computer, or the like. In detail, the electronic device comprises at least, connected by a bus 41: a memory 42 and a processor 43, wherein the memory 42 is used for storing a computer program, and the processor 43 is used for executing the computer program stored in the memory 42 to execute all or part of the steps in the foregoing method embodiments.
The above-mentioned system bus may be a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, or the like. The system bus may be divided into an address bus, a data bus, a control bus, and the like. For ease of illustration, only one thick line is shown, but this does not mean that there is only one bus or one type of bus. The communication interface is used for realizing communication between the database access device and other equipment (such as a client, a read-write library and a read-only library). The Memory may include a Random Access Memory (RAM), and may further include a non-volatile Memory (non-volatile Memory), such as at least one disk Memory.
The Processor may be a general-purpose Processor, and includes a Central Processing Unit (CPU), a Network Processor (NP), and the like; the Integrated Circuit may also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA) or other Programmable logic device, a discrete Gate or transistor logic device, or a discrete hardware component.
In summary, the house type graph scale detection method, the house type graph scale detection device and the electronic equipment based on deep learning provided by the invention can improve the recall rate of the model through text recognition and pixel-level template matching positioning, realize self-verification of the detection result through a statistical learning method, and improve the precision rate of scale detection, thereby improving the design efficiency of a home decoration designer and accelerating the development of digital construction of the home decoration industry. Therefore, the invention effectively overcomes various defects in the prior art and has high industrial utilization value.
Finally, it should be noted that: the above embodiments are only used to illustrate the technical solution of the present invention, and not to limit the same; while the invention has been described in detail and with reference to the foregoing embodiments, it will be understood by those skilled in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some or all of the technical features may be equivalently replaced; and the modifications or the substitutions do not make the essence of the corresponding technical solutions depart from the scope of the technical solutions of the embodiments of the present invention.
Claims (10)
1. A house type graph scale detection method based on deep learning is characterized by comprising the following steps:
detecting the position of a ruler area of the house pattern;
acquiring a physical size area position in the scale area;
splitting and identifying physical dimension values within the physical dimension region;
constructing a scale end point template pair to be matched with the scale region, acquiring the position of the scale end point template pair when the matching degree exceeds a preset threshold value, and determining the position of the scale end point pair, wherein the scale end point template pair consists of a pair of opposite scale end point templates, and a pair of scale end points form a section of scale;
determining the pixel number of the scale according to the position of the scale end point pair;
pairing the physical size and the scale and determining the house type graph scale according to the number of pixels of the scale;
and verifying the scale accuracy of the house type graph.
2. The method of claim 1, further comprising:
converting the floor-type map into a gray-scale map and calculating a pixel histogram of the gray-scale map;
determining the gray-scale image pixel conversion threshold value according to the pixel histogram;
and converting the gray-scale image into a black-white image according to the pixel conversion threshold value.
3. The method of claim 2, further comprising:
detecting a rectangular area where the house type graph is located according to the trained house type graph detection model and acquiring the vertex coordinates of the area;
expanding the rectangular area where the house type graph is located to the peripheral area according to the vertex coordinates;
performing expansion corrosion treatment on the expansion area, wherein the expansion area after treatment is the scale area;
and acquiring the coordinate position of the scale area.
4. The method of claim 1 or 3, further comprising:
detecting the physical size in the scale area and positioning to obtain the position coordinate of the physical size;
resolving numbers in the physical dimensions by erosion dilation;
and predicting the numerical corresponding value in the physical dimension according to a pre-established numerical prediction model.
5. The method of claim 1 or 3, further comprising:
constructing a corresponding ruler endpoint template pair according to the type of the end point of the house type icon ruler;
traversing the scale region of the house-type figure through the scale endpoint template, and calculating the matching degree of the scale endpoint template and the scale region according to the number of the overlapped pixel values of the scale endpoint template and the scale region of the house-type figure.
6. The method of claim 1, further comprising:
matching the physical size value with the scale according to the position of the connecting line of the physical size area position and the scale end point pair;
determining a scale of a corresponding scale according to the physical size and the scale pixel number which are successfully matched;
and determining the scale of the house type graph according to the scale value with the deviation value smaller than the preset threshold value.
7. The method of claim 1, further comprising:
comparing and determining the product of the house type graph scale and the scale pixel number with the deviation value of the scale physical dimension value;
and when the deviation value of the scale is smaller than the preset deviation value threshold, the scale accuracy of the house type graph meets the standard, otherwise, the scale is detected again.
8. A house type graph scale detection device based on deep learning is characterized in that the device comprises:
the detection module is used for detecting and acquiring the position of the scale area, the position of the physical size area and the position of the scale end point pair;
the text recognition module is used for processing and predicting the physical size value in the physical size area;
the construction module is used for constructing the scale endpoint template pair according to the scale endpoint types;
the matching module is used for matching the scale endpoint template pair with the scale area;
the calculation module is used for calculating the scale of the house type graph;
and the verification module is used for verifying the accuracy of the scale of the house type graph.
9. A computer-readable storage medium, in which a computer program is stored, which, when loaded and executed by a processor, implements a deep learning-based house type graph scale detection method according to any one of claims 1 to 7.
10. An electronic device, comprising: a processor, a memory; wherein the content of the first and second substances,
the memory is used for storing a computer program;
the processor is used for loading and executing the computer program to enable the electronic equipment to execute the deep learning based house type graph scale detection method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110656784.3A CN113392455B (en) | 2021-06-11 | House pattern scale detection method and device based on deep learning and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110656784.3A CN113392455B (en) | 2021-06-11 | House pattern scale detection method and device based on deep learning and electronic equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113392455A true CN113392455A (en) | 2021-09-14 |
| CN113392455B CN113392455B (en) | 2024-06-25 |
Family
ID=
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114882306A (en) * | 2022-04-06 | 2022-08-09 | 国家基础地理信息中心 | Topographic map scale identification method and device, storage medium and electronic equipment |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108763606A (en) * | 2018-03-12 | 2018-11-06 | 江苏艾佳家居用品有限公司 | A kind of floor plan element extraction method and system based on machine vision |
| CN110414477A (en) * | 2019-08-06 | 2019-11-05 | 广东三维家信息科技有限公司 | Image scale detection method and device |
| CN111222508A (en) * | 2020-01-13 | 2020-06-02 | 江苏艾佳家居用品有限公司 | ROI-based house type graph scale identification method and device and computer equipment |
| CN111291649A (en) * | 2020-01-20 | 2020-06-16 | 广东三维家信息科技有限公司 | Image recognition method and device and electronic equipment |
| CN111311671A (en) * | 2020-05-12 | 2020-06-19 | 创新奇智(南京)科技有限公司 | Workpiece measuring method and device, electronic equipment and storage medium |
| CN111611643A (en) * | 2020-05-27 | 2020-09-01 | 电子科技大学中山学院 | Family type vectorization data obtaining method and device, electronic equipment and storage medium |
| CN112052489A (en) * | 2019-07-22 | 2020-12-08 | 江苏艾佳家居用品有限公司 | Method and system for generating house type graph |
| CN112257506A (en) * | 2020-09-21 | 2021-01-22 | 北京豆牛网络科技有限公司 | Fruit and vegetable size identification method and device, electronic equipment and computer readable medium |
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108763606A (en) * | 2018-03-12 | 2018-11-06 | 江苏艾佳家居用品有限公司 | A kind of floor plan element extraction method and system based on machine vision |
| CN112052489A (en) * | 2019-07-22 | 2020-12-08 | 江苏艾佳家居用品有限公司 | Method and system for generating house type graph |
| CN110414477A (en) * | 2019-08-06 | 2019-11-05 | 广东三维家信息科技有限公司 | Image scale detection method and device |
| CN111222508A (en) * | 2020-01-13 | 2020-06-02 | 江苏艾佳家居用品有限公司 | ROI-based house type graph scale identification method and device and computer equipment |
| CN111291649A (en) * | 2020-01-20 | 2020-06-16 | 广东三维家信息科技有限公司 | Image recognition method and device and electronic equipment |
| CN111311671A (en) * | 2020-05-12 | 2020-06-19 | 创新奇智(南京)科技有限公司 | Workpiece measuring method and device, electronic equipment and storage medium |
| CN111611643A (en) * | 2020-05-27 | 2020-09-01 | 电子科技大学中山学院 | Family type vectorization data obtaining method and device, electronic equipment and storage medium |
| CN112257506A (en) * | 2020-09-21 | 2021-01-22 | 北京豆牛网络科技有限公司 | Fruit and vegetable size identification method and device, electronic equipment and computer readable medium |
Non-Patent Citations (2)
| Title |
|---|
| 马博: "基于深度学习的户型图结构提取方法", 《中国优秀硕士学位论文全文数据库工程科技Ⅱ辑》, 1 May 2020 (2020-05-01), pages 038 - 138 * |
| 高智慧 等: "基于ArcGIS的可拓住宅建筑设计数据库构建研究", 智能系统学报, no. 02, pages 1 - 7 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114882306A (en) * | 2022-04-06 | 2022-08-09 | 国家基础地理信息中心 | Topographic map scale identification method and device, storage medium and electronic equipment |
| CN114882306B (en) * | 2022-04-06 | 2023-08-18 | 国家基础地理信息中心 | Topography scale identification method and device, storage medium and electronic equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110060237B (en) | Fault detection method, device, equipment and system | |
| CN108710866B (en) | Chinese character model training method, chinese character recognition method, device, equipment and medium | |
| CN108920580B (en) | Image matching method, device, storage medium and terminal | |
| KR20210110823A (en) | Image recognition method, training method of recognition model, and related devices and devices | |
| CN111860502A (en) | Picture table identification method and device, electronic equipment and storage medium | |
| CN109389110B (en) | Region determination method and device | |
| CN108734161B (en) | Method, device and equipment for identifying prefix number area and storage medium | |
| CN110807404A (en) | Form line detection method, device, terminal and storage medium based on deep learning | |
| CN111340023A (en) | Text recognition method and device, electronic equipment and storage medium | |
| CN116030237A (en) | Industrial defect detection method and device, electronic equipment and storage medium | |
| CN115273115A (en) | Document element labeling method and device, electronic equipment and storage medium | |
| CN112634201A (en) | Target detection method and device and electronic equipment | |
| CN112052702A (en) | Method and device for identifying two-dimensional code | |
| CN113129298B (en) | Method for identifying definition of text image | |
| CN117474918B (en) | Abnormality detection method and device, electronic device, and storage medium | |
| CN110751156A (en) | Method, system, device and medium for table line bulk interference removal | |
| CN111192279A (en) | Object segmentation method based on edge detection, electronic terminal and storage medium | |
| CN111340788B (en) | Hardware Trojan horse layout detection method and device, electronic equipment and readable storage medium | |
| CN112784494A (en) | Training method of false positive recognition model, target recognition method and device | |
| CN109726722B (en) | Character segmentation method and device | |
| CN114973300B (en) | Component type identification method and device, electronic equipment and storage medium | |
| CN113392455B (en) | House pattern scale detection method and device based on deep learning and electronic equipment | |
| CN110751623A (en) | Joint feature-based defect detection method, device, equipment and storage medium | |
| CN113392455A (en) | House type graph scale detection method and device based on deep learning and electronic equipment | |
| CN115272682A (en) | Target object detection method, target detection model training method and electronic equipment |
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 | ||
| GR01 | Patent grant |