CN111028143A - Design method for invariant features of different scale transformation of image - Google Patents
Design method for invariant features of different scale transformation of image Download PDFInfo
- Publication number
- CN111028143A CN111028143A CN201911231422.9A CN201911231422A CN111028143A CN 111028143 A CN111028143 A CN 111028143A CN 201911231422 A CN201911231422 A CN 201911231422A CN 111028143 A CN111028143 A CN 111028143A
- Authority
- CN
- China
- Prior art keywords
- image
- gaussian
- invariant
- moment
- scale
- 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
- 238000000034 method Methods 0.000 title claims abstract description 32
- 230000009466 transformation Effects 0.000 title claims abstract description 29
- 238000013461 design Methods 0.000 title claims abstract description 18
- 230000008901 benefit Effects 0.000 abstract description 10
- 238000010586 diagram Methods 0.000 description 8
- 238000000844 transformation Methods 0.000 description 8
- 238000004590 computer program Methods 0.000 description 7
- 230000008569 process Effects 0.000 description 6
- 238000012545 processing Methods 0.000 description 6
- 238000012360 testing method Methods 0.000 description 5
- 230000006870 function Effects 0.000 description 4
- 238000011160 research Methods 0.000 description 4
- 230000009471 action Effects 0.000 description 3
- 238000004422 calculation algorithm Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 2
- 238000003384 imaging method Methods 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- 210000004556 brain Anatomy 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 230000001149 cognitive effect Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000036403 neuro physiology Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 238000010200 validation analysis Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformations in the plane of the image
- G06T3/18—Image warping, e.g. rearranging pixels individually
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The embodiment of the invention provides a design method for invariant features of image different-scale transformation, and relates to the field of computer vision. The method comprises the following steps: calculating the geometric center distance of the image to be processed; obtaining the weights of the scale parameters of different dimensions according to the geometric center distance; acquiring new scale parameters of different dimensions according to the weight; calculating Gaussian-Hemrite moment according to the new scale parameters; the non-uniform scale invariant is obtained by any two updated Gaussian-Hemrite moments. The embodiment of the invention designs the invariant moment based on the orthogonal Gaussian-Hermite moment ratio, realizes the design of the steady characteristics under the condition that the target is subjected to non-uniform ratio transformation, and has the advantages of excellent expansibility and good numerical stability.
Description
Technical Field
The invention relates to the field of computer vision, in particular to a design method for invariant features of different scale transformations of an image.
Background
Computer vision is a question of how to use cameras and computers to obtain the data and information that we need about the object to be photographed. Pictorially, it is to install eyes (camera) and brain (algorithm) to a computer to make the computer sense the environment. Computer vision is a challenging important research area in both engineering and science. Computer vision is a comprehensive discipline that has attracted researchers from various disciplines to participate in its research. Including computer science and engineering, signal processing, physics, applied mathematics and statistics, neurophysiology and cognitive sciences, and the like.
In the field of computer vision, the most basic requirement for accurately identifying a target is that the description of target characteristics must have invariance to the scale transformation of the target. The existing design method of scale invariant features mainly aims at that a target has consistent scale transformation in the x and y directions or the x, y and z directions, such as geometric scale invariant moment, Legendre and Krawtchouk invariant moment. Some comprehensive feature design methods can realize different scale transformation invariant features, such as normalization of images, SIFT correlation algorithm, and Tchebichef scale invariant. However, the complexity of implementation of the comprehensive algorithms and features is high, and the real-time performance is difficult to meet. Furthermore, good numerical stability of the features is a potential requirement. The invariant features designed for different scaling should have scalability and high numerical stability. The existing geometric scale invariant moment is poor in numerical stability, so that the existing geometric scale invariant moment is rarely adopted in engineering practice.
At present, the studies to design a proportional invariant based on the Gaussian-Hermite moment are the following: [1] yang, J.Kostkova, J.Fluser, T.Suk, Scale Invariants from Gaussian-HermitiMemes, Signal Processing,132(2017)77-84.
The document only discusses the design of Gaussian-Hermite ratio invariant moment under the condition that the same ratio transformation occurs in the x and y directions. The specific realization is to adopt 0-order geometric moment m of the image00The weight as a scale parameter:
the method proposed in document [1] has no invariance when the two-dimensional image f (x, y) or the three-dimensional image f (x, y, z) has different scale transformations in all directions, and the invention can ensure that the calculated Gaussian-Hermite moment characteristic has invariance in the situation.
The embodiment of the invention provides a design method for the characteristic of invariant conversion of images in different proportions, and the method is based on a Gaussian-Hermite orthogonal moment, namely the characteristic of non-uniform-scale Gaussian-Hermite invariant moment, and has the advantages of convenience in calculation, excellent expansibility and good numerical stability.
Disclosure of Invention
In view of the above, embodiments of the present invention are proposed to provide a design method for different scaling invariant features of an image that overcomes or at least partially solves the above mentioned problems.
In order to solve the above problem, an embodiment of the present invention discloses a method for designing an invariant feature for different scaling transformations of an image, including:
calculating the geometric center distance of the image to be processed;
obtaining the weight values of the scale parameters of different dimensions according to the geometric center distance;
acquiring new scale parameters of different dimensions according to the weight;
calculating a new Gaussian-Hemrite moment according to the new scale parameters;
the non-uniform scale invariant is obtained by any two updated Gaussian-Hemrite moments.
Preferably, the step of calculating a new Gaussian-Hemrite moment according to the new scale parameters further comprises: and acquiring the barycentric coordinates of the image to be processed.
The embodiment of the invention has the following advantages:
the embodiment of the invention aims at the stable orthogonal moment characteristic under the condition that two-dimensional and three-dimensional images have uneven proportional transformation. Scaling is one of the most fundamental, important geometric transformations in computer vision. The non-uniform scaling transformation does not belong to rigid body transformation, and is ubiquitous in the sensor imaging process. Designing features that are invariant to non-uniform scaling is therefore a fundamental matter of research in the field of computer vision. The embodiment of the invention designs the invariant moment based on the orthogonal Gaussian-Hermite moment ratio, and realizes the characteristic design problem under the condition that the target generates the non-uniform ratio transformation. The characteristic has the advantages of convenient calculation, excellent expansibility and good numerical stability.
The Gaussian-Hermite proportion invariant moment provided by the invention inherits the advantage of good numerical stability of the Gaussian-Hermite orthogonal moment. Meanwhile, the invariant moment of the Gaussian-Hermite proportion provided by the invention can be expanded to a high order and a multidimensional signal. The advantage is convenient for popularization in engineering practice, and is a technology which can be directly applied to engineering.
Drawings
FIG. 1 is a flow chart of the steps of a method for designing an invariant feature for different scaling of an image according to the present invention.
FIG. 2 is a three-dimensional image sample of varying dimensions used to test the method of the present invention.
Fig. 3 is a sample of actual photographs with different dimensions used to test the method of the invention.
Fig. 4 is a graph of invariant moment values for different sizes of actual photographic samples.
Detailed Description
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in further detail below.
Example one
Referring to fig. 1, a flowchart illustrating steps of a method for designing an invariant feature for different scaling transformations of an image according to the present invention is shown, which may specifically include the following steps:
and step S101, calculating the geometric center distance of the image to be processed.
If the image to be processed is a two-dimensional image f (x, y), the geometric center distance mup,qCalculated according to equation (1).
Wherein p and q are non-negative integers. Correspondingly, if the image to be processed is a three-dimensional image f (x, y, z), the geometric center distance μp,q,rCalculated according to equation (2).
Wherein p, q and r are non-negative integers. Preferably, the image to be processed according to the present invention includes, but is not limited to, a two-dimensional image and a three-dimensional image.
And S102, acquiring the weight values of the scale parameters with different dimensions according to the geometric center distance.
If the image to be processed is a two-dimensional image f (x, y), the horizontal direction scale parameter sigmaxAnd the vertical direction sigmayThe weights of (a) are calculated by the formula (3), respectively.
Wherein images of different sizes are computed to obtain mu20,μ02And mu00The values of (c) are different. Preferably, if the image to be processed is a three-dimensional image f (x, y, z), calculating a weight of each dimension (x, y, z) scale parameter of the three-dimensional image, as shown in formula (4):
and S103, acquiring new scale parameters with different dimensions according to the weight.
Acquiring the weight values of the scale parameters of different dimensions according to the step S102, and acquiring new scale parameters of Gaussian-Hermite moment x and y directions for the two-dimensional image:
wherein σxAnd σyIs a scale factor of Gaussian-Hermite moment and is determined according to the image to be processed. Preferably, if the processed image is a three-dimensional image f (x, y, z), new scale parameters in x, y and z directions are acquired:
and step S104, calculating a new Gaussian-Hemrite moment according to the new scale parameters.
The image to be processed is a two-dimensional image f (x, y), and the (p, q) -order Gaussian-Hermite moment of the image to be processed is defined as shown in the following formula (7).
Wherein (x) in the formula (7)c,yc) Is the barycentric coordinates of the image, calculated by equation (8); hp(x) Is a Hermite polynomial; sigmaxAnd σyAre scale parameters in the x and y directions; Ω is the domain of definition of the image. Accordingly, if the processed image is a three-dimensional image f (x, y, z), the three-dimensional Gaussian-Hermite moment is as shown in equation (9).
If the image to be processed is a two-dimensional image f (x, y), the image in the formula (5) is processed
And
respectively replace sigma in formula (7)xAnd σyA new set of (p, q) order Gaussian-Hermite moments is calculated as shown in equation (10).
Correspondingly, if the image to be processed is a three-dimensional image f (x, y, z), new scale parameters in the x, y and z directions in the formula (6) are substituted into the formula (9), and a new (p, q, r) order Gaussian-Hermite moment group is calculated as shown in the formula (11).
Step S105: the non-uniform scale invariant is obtained by any two updated Gaussian-Hemrite moments.
Through any two updated Gaussian-Hermite moments, an invariant with invariance to the non-uniform scale transformation can be obtained, and as shown in a formula (12), the invariant of the image f (x, y) to be processed is obtained. By using
The normal parameters are able to eliminate the variation of the new Gaussian-Hermite moment scaling factor.
The normal parameter in the formula (12) may be
But are not limited to
If the processed image is a three-dimensional image, acquiring a non-uniform scale invariant through the ratio of any two new three-dimensional Gaussian-Hermite moments, such as a formula (13), and acquiring the invariant of the image f (x, y, z) to be processed.
Phi obtained from formula (12) and formula (13), respectivelyp,qAnd phip,q,rNamely invariant of the image to be processed which is independent of different dimensional transformations of the respective dimensions.
According to the embodiment of the invention, the unchanged characteristics of different scale changes of the measurement sample after the scale changes are carried out on the plurality of three-dimensional images and the sample image of the photo with different sizes are obtained.
(1) Verification of measurement sample after scale change of three-dimensional image
For public three-dimensional body database Prin50 different three-dimensional forms were selected from ceton Shape Benchmark. Random x, y and z-dimension transformations are performed on each shape, and three scaled test samples are shown in fig. 2 for one of the samples. The three test specimens have different geometric moments m of order 0000I.e. the value is different for the same object after different scaling, which means that the three samples have different sizes. The following Gaussian-Hermite scale invariant moment is calculated for the transformed shape:
V＝[Φ200,Φ111,Φ201,Φ202,Φ112,Φ131,Φ005,Φ222,Φ600,Φ411,Φ402,Φ343]chinese character of 'ji' (14)
The ratio invariance included in equation (14) was calculated for each of 50 samples of the three-dimensional shape.
The following error measures were used:
wherein the content of the first and second substances,
and
respectively, the standard deviation and the mean of the ith moment of inertia. The test results were as follows: a total of 900 torque transmitting elements, of which 814 have an error of less than 10%.
(2) Sample validation of different size photographs
As fig. 3 shows the actual photographs scanned, the photographs themselves have different sizes. For each picture, calculating two-dimensional proportional invariant moment phi20，Φ21，Φ03，Φ40，Φ22，Φ50And phi23. Fig. 4 shows the two-dimensional scale invariant moment values corresponding to the respective photographs. As can be seen from FIG. 4, the same invariant moment values corresponding to different photographs are substantially the same, Φ201.7453%;Φ213.0893%; phi032.2212%; phi401.7558%; phi221.7716%; phi505.4233%; phi23The error of (2) is 3.0303%. Actual data show that the non-uniform ratio transformation Gaussian-Hermite invariant moment provided by the invention has invariance to the non-uniform ratio transformation and has better stability in numerical value.
The embodiment of the invention aims at the design of the stable orthogonal moment characteristic under the condition that two-dimensional and three-dimensional images have uneven proportional transformation. Scaling is one of the most fundamental, important geometric transformations in machine vision. The non-uniform scaling transformation does not belong to rigid body transformation, and is ubiquitous in the sensor imaging process. Designing features that are invariant to non-uniform scaling is therefore a fundamental element in the field of computer vision research. The invention designs the proportional invariant moment based on the orthogonal Gaussian-Hermite moment, and realizes the characteristic design problem under the condition that the target generates non-uniform proportional transformation. The characteristic has the advantages of convenient calculation, excellent expansibility and good numerical stability.
The Gaussian-Hermite proportion invariant moment provided by the embodiment of the invention inherits the advantage of good numerical stability of the Gaussian-Hermite orthogonal moment. Meanwhile, the invariant moment of the Gaussian-Hermite proportion provided by the invention can be expanded to a high order and a multidimensional signal. The advantage is convenient for popularization in engineering practice, and is a technology which can be directly applied to engineering.
It should be noted that, for simplicity of description, the method embodiments are described as a series of acts or combination of acts, but those skilled in the art will recognize that the present invention is not limited by the illustrated order of acts, as some steps may occur in other orders or concurrently in accordance with the embodiments of the present invention. Further, those skilled in the art will appreciate that the embodiments described in the specification are presently preferred and that no particular act is required to implement the invention.
The embodiments in the present specification are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other.
As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, apparatus, or computer program product. Accordingly, embodiments of the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, embodiments of the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
Embodiments of the present invention are described with reference to flowchart illustrations and/or block diagrams of methods, terminal devices (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing terminal to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing terminal, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing terminal to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing terminal to cause a series of operational steps to be performed on the computer or other programmable terminal to produce a computer implemented process such that the instructions which execute on the computer or other programmable terminal provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
While preferred embodiments of the present invention have been described, additional variations and modifications of these embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. Therefore, it is intended that the appended claims be interpreted as including preferred embodiments and all such alterations and modifications as fall within the scope of the embodiments of the invention.
Finally, it should also be noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or terminal that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or terminal. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, method, article, or terminal that comprises the element.
The present invention provides a method for designing an invariant feature of image transformation in different scales, and a specific example is applied in the text to explain the principle and the implementation of the present invention, and the description of the above embodiment is only used to help understanding the method and the core idea of the present invention; meanwhile, for a person skilled in the art, according to the idea of the present invention, there may be variations in the specific embodiments and the application scope, and in summary, the content of the present specification should not be construed as a limitation to the present invention.
Claims (2)
1. A design method for the feature that the image has different dimensions and is not changed in the scaling transformation is characterized by comprising the following steps:
calculating the geometric center distance of the image to be processed;
obtaining the weights of the scale parameters of different dimensions according to the geometric center distance;
acquiring new scale parameters of different dimensions according to the weight;
calculating a new Gaussian-Hemrite moment according to the new scale parameter;
the non-uniform scale invariant is obtained by any two updated Gaussian-Hemrite moments.
2. The method of claim 1, wherein the step of calculating a new Gaussian-Hemrite moment from the new scale parameters is preceded by the step of: and acquiring the barycentric coordinates of the image to be processed.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911231422.9A CN111028143B (en) | 2019-12-05 | 2019-12-05 | Design method for invariant features of different scale transformation of image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911231422.9A CN111028143B (en) | 2019-12-05 | 2019-12-05 | Design method for invariant features of different scale transformation of image |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111028143A true CN111028143A (en) | 2020-04-17 |
| CN111028143B CN111028143B (en) | 2022-11-11 |
Family
ID=70208030
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911231422.9A Active CN111028143B (en) | 2019-12-05 | 2019-12-05 | Design method for invariant features of different scale transformation of image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111028143B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113888721A (en) * | 2021-10-08 | 2022-01-04 | 西北工业大学 | Three-dimensional image feature design method based on integer-fractional order orthogonal polynomial |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130061184A1 (en) * | 2011-09-02 | 2013-03-07 | International Business Machines Corporation | Automated lithographic hot spot detection employing unsupervised topological image categorization |
| CN103630899A (en) * | 2013-03-29 | 2014-03-12 | 中国科学院电子学研究所 | Method for high-resolution radar compressed sensing imaging of moving object on ground |
| CN108595649A (en) * | 2018-04-27 | 2018-09-28 | 郑州轻工业学院 | The textile image search method of local invariant textural characteristics based on geometry |
| CN109871625A (en) * | 2019-02-26 | 2019-06-11 | 西南交通大学 | Non-gaussian wind pressure analogy method based on Johnson transformation |
-
2019
- 2019-12-05 CN CN201911231422.9A patent/CN111028143B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130061184A1 (en) * | 2011-09-02 | 2013-03-07 | International Business Machines Corporation | Automated lithographic hot spot detection employing unsupervised topological image categorization |
| CN103630899A (en) * | 2013-03-29 | 2014-03-12 | 中国科学院电子学研究所 | Method for high-resolution radar compressed sensing imaging of moving object on ground |
| CN108595649A (en) * | 2018-04-27 | 2018-09-28 | 郑州轻工业学院 | The textile image search method of local invariant textural characteristics based on geometry |
| CN109871625A (en) * | 2019-02-26 | 2019-06-11 | 西南交通大学 | Non-gaussian wind pressure analogy method based on Johnson transformation |
Non-Patent Citations (2)
| Title |
|---|
| BO YANG等: "《Scale invariants from Gaussian–Hermite moments》", 《HTTP://DX.DOI.ORG/10.1016/J.SIGPRO.2016.09.013》 * |
| 谢永华: "基于不变矩特征木材纹理的研究", 《林业机械与木工设备》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113888721A (en) * | 2021-10-08 | 2022-01-04 | 西北工业大学 | Three-dimensional image feature design method based on integer-fractional order orthogonal polynomial |
| CN113888721B (en) * | 2021-10-08 | 2022-09-16 | 西北工业大学 | Three-dimensional image feature design method based on integer-fractional order orthogonal polynomial |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111028143B (en) | 2022-11-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Goshtasby | Image registration: Principles, tools and methods | |
| Bovik | The essential guide to image processing | |
| Rahtu et al. | Affine invariant pattern recognition using multiscale autoconvolution | |
| Chaudhuri et al. | Depth from defocus: a real aperture imaging approach | |
| CN102833460B (en) | Image processing method, image processing device and scanner | |
| DE60036780T2 (en) | A method of comparing a two-dimensional image to one of a plurality of three-dimensional candidate models stored in a database | |
| Ali et al. | Analysis of blur measure operators for single image blur segmentation | |
| Manzanera et al. | Line and circle detection using dense one-to-one Hough transforms on greyscale images | |
| Tafti et al. | A comparative study on the application of SIFT, SURF, BRIEF and ORB for 3D surface reconstruction of electron microscopy images | |
| CN111340077B (en) | Attention mechanism-based disparity map acquisition method and device | |
| Griffin | The second order local-image-structure solid | |
| CN114925748B (en) | Model training and modal information prediction method, related device, equipment and medium | |
| CN112435193A (en) | Method and device for denoising point cloud data, storage medium and electronic equipment | |
| CN111179270A (en) | Image co-segmentation method and device based on attention mechanism | |
| JP5704909B2 (en) | Attention area detection method, attention area detection apparatus, and program | |
| Frantc et al. | Machine learning approach for objective inpainting quality assessment | |
| CN111028143B (en) | Design method for invariant features of different scale transformation of image | |
| JP6645442B2 (en) | Information processing apparatus, information processing method, and program | |
| US8208731B2 (en) | Image descriptor quantization | |
| CN110793441A (en) | High-precision object geometric dimension measuring method and device | |
| US20080267506A1 (en) | Interest point detection | |
| Ren et al. | A non-contact sleep posture sensing strategy considering three dimensional human body models | |
| Wu et al. | Screen content image quality assessment based on the most preferred structure feature | |
| JP2020047010A (en) | Information estimation device and information estimation method | |
| JP6202938B2 (en) | Image recognition apparatus and image recognition method |
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 | ||
| GR01 | Patent grant |