CN109934093A - A kind of method, computer-readable medium and identifying system identifying commodity on shelf - Google Patents
A kind of method, computer-readable medium and identifying system identifying commodity on shelf Download PDFInfo
- Publication number
- CN109934093A CN109934093A CN201910055003.8A CN201910055003A CN109934093A CN 109934093 A CN109934093 A CN 109934093A CN 201910055003 A CN201910055003 A CN 201910055003A CN 109934093 A CN109934093 A CN 109934093A
- Authority
- CN
- China
- Prior art keywords
- commodity
- shelf
- picture
- product features
- coordinate system
- 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
Abstract
The present invention relates to a kind of methods for identifying commodity on shelf, include the following steps, step S1: the commodity on shelf are shot at least two pictures;Step S2: the product features in detection picture;Step S3: establishing coordinate system, obtains commodity phase adjacency pair, and be labeled to phase adjacency pair;Step S4: identical mark in adjacent two picture is subjected to overlapping mapping, all pictures are spliced into a picture, to identify commodity.The present invention also provides a kind of computer-readable mediums.The present invention also provides a kind of identifying systems.
Description
[technical field]
The present invention relates to unmanned retail domain, provide a kind of method for identifying commodity on shelf, computer-readable medium and
Identifying system.
[background technique]
The existing unmanned shop method of commodity on shelf of making an inventory generally manually makes an inventory or is made an inventory by shooting picture,
The main means that shooting picture is made an inventory are that the commodity in shelf are shot for multiple pictures for having coincidence mutually, are calculated by surf etc.
Method finds the product features in picture, and removes the product features of intersection in adjacent picture, finally splices to picture
Realization makes an inventory to commodity on shelf.When being placed with a large amount of similar or identical commodity in shelf, then it is difficult to distinguish adjacent
The part being overlapped in picture causes finally to splice error.
[summary of the invention]
Of the existing technology to overcome the problems, such as, the present invention provides a kind of method, computer-readable for identifying commodity on shelf
Medium and identifying system.
The scheme that the present invention solves technical problem is to provide a kind of method for identifying commodity on shelf, for identification on shelf
Commodity, this approach includes the following steps, step S1: the commodity on shelf are shot at least two pictures;Step S2: detection figure
Product features in piece;Step S3: establishing coordinate system, obtains commodity phase adjacency pair, and be labeled to phase adjacency pair;Step S4: will
Identical mark carries out overlapping mapping in adjacent two picture, and all pictures are spliced into a picture, to identify quotient
Product.
Preferably, step S3 further comprises the steps of: step S31: establishing coordinate system;Step S32: in a coordinate system according to commodity
Position, determine the positional relationship between adjacent commodity;Step S33: in conjunction with the positional relationship definition between product features and commodity
Phase adjacency pair, and using between commodity positional relationship and product features as the adjacent pairs of mark.
Preferably, the phase adjacency pair is three commodity adjacent each other in coordinate system.
Preferably, high-level semantic is used to select the not factors such as light irradiation angle influence after detecting the product features in picture
Mark of the product features as commodity.
Preferably, the product features include the external appearance characteristic and position feature of commodity.
Preferably, the external appearance characteristic includes the shape, color and pattern of commodity, and the position feature is commodity in coordinate
Coordinate in system.
Preferably, before the commodity on shelf being shot into picture, first marked articles information, merchandise news includes commodity
External appearance characteristic, title and unit price.
Preferably, after the commodity identified, distribution situation, institute accounting of the commodity on shelf are obtained according to merchandise news
Example and commodity total price.
The present invention also provides a kind of computer-readable mediums, it is characterised in that: is stored in the computer-readable medium
Computer program, wherein the computer program is arranged to the method that above-mentioned identification commodity on shelf is executed when operation.
The present invention also provides a kind of identifying systems, it is characterised in that: and the commodity on shelf check system includes shooting module,
For shooting the commodity on shelf at picture;Detection module, for detecting the product features in picture;Coding module is used for
Coordinate system is established, obtains commodity phase adjacency pair, and be labeled to phase adjacency pair;Module is integrated, it will be identical in adjacent two picture
Mark carries out overlapping mapping, all pictures is spliced into a picture, to identify commodity.
Compared with prior art, the method for identification commodity on shelf of the invention has the advantage that
1. adjacent pairs include lap, uses phase adjacency pair as the feature of a commodity, receptive field can be expanded, avoided
Because when only using a product features and being identified, lead in same picture that the same category product features are excessively similar to be caused
Erroneous matching.
2. the influence of the product features not factors such as light irradiation angle of high-level semantic selection passes through the commodity that high-level semantic selects
Feature can more accurately identify its corresponding commodity.
3. the mark in all pictures is sequentially mapped in same plane, repeating part using overlapping mapping, then from
Distribution situation, proportion and commodity total price of the commodity on shelf can be intuitively obtained in the picture of splicing.
[Detailed description of the invention]
Fig. 1 is the method flow schematic diagram of first embodiment of the invention identification commodity on shelf.
Fig. 2 is the flow diagram of step S3 in the method for first embodiment of the invention identification commodity on shelf.
Fig. 3 A is the mark schematic diagram of the method for first embodiment of the invention identification commodity on shelf.
Fig. 3 B is the method for first embodiment of the invention identification commodity on shelf according to the schematic diagram of mark mapping.
Fig. 4 is the module diagram of third embodiment of the invention identifying system.
Description of symbols: 1, identifying system;11, shooting module;12, detection module;13, coding module;14, mould is integrated
Block;131, modeling module;132, analysis module;133, labeling module.
[specific embodiment]
In order to make the purpose of the present invention, technical solution and advantage are more clearly understood, below in conjunction with attached drawing and embodiment,
The present invention will be described in further detail.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention,
It is not intended to limit the present invention.
Referring to Fig. 1, first embodiment of the invention provides a kind of method for identifying commodity on shelf, for identification on shelf
Commodity, this approach includes the following steps,
Step S1: the commodity on shelf are shot at least two pictures;
Step S2: the product features in detection picture;
Step S3: establishing coordinate system, obtains commodity phase adjacency pair, and be labeled to phase adjacency pair;
Step S4: identical mark in adjacent two picture is subjected to overlapping mapping, all pictures are spliced into one
Picture, to identify commodity.
It first can not be by goods by the shooting of single due to shelf length issue at picture by the commodity shooting on shelf
All commodity on frame all take in a picture, therefore at least will be by shooting twice, could be by the whole on shelf
Commodity are all filmed.There are commodity to omit and influence recognition result when shooting in order to prevent, so can be selected when shooting by phase
Two adjacent pictures partially overlap shooting;Then the product features in picture are detected according to the picture of shooting, detected in picture
High-level semantic is used to select mark of the product features of the factors such as light irradiation angle influence as commodity after product features;Then
Coordinate system is established, commodity phase adjacency pair is obtained from coordinate system, and mark phase adjacency pair using semantic coding, it is each adjacent to distinguish
It is right;Mark adjacent pairs of in all pictures is finally mapped to same plane, identical mark carries out weight in adjacent two picture
All pictures are spliced into a picture, the commodity on shelf are gone out according to the picture recognition of splicing by folded mapping.
Commodity phase adjacency pair is at least two commodity adjacent each other in coordinate system, can be two, three, four or more
It is a, it is preferable that phase adjacency pair is three commodity adjacent each other in pixel coordinate system.Commodity phase adjacency pair can be in image edge area
Domain obtains, i.e., in the position acquisition far from coordinate origin;It arbitrary region can also be obtained in picture, i.e., in a coordinate system
Arbitrary region obtains.
It is appreciated that the sequence that commodity are put along shelf by the picture of shooting successively sorts, to obtain adjacent two
Picture.Identical mark is duplicate part in adjacent two picture in adjacent two picture, to avoid repeating part from influencing
Count shelf on commodity as a result, therefore to carry out overlapping mapping to identical mark.Overlapping is mapped as two identical marks
Note mapping identical position, therefore two marks only display one after overlapping mapping in the same plane.Detect picture
The method of middle product features is that detected the product features in each picture using neural network.On shooting shelf
Commodity can be artificial shooting, be also possible to electronic monitoring shooting.
It is appreciated that product features be detected by neural network, product features include commodity external appearance characteristic and
Position feature, external appearance characteristic include shape, color and pattern, and position feature is the coordinate of commodity in a coordinate system.In shelf
It is placed with one bottle of black tea, shape, color and the pattern having in itself of black tea are external appearance characteristic, the coordinate of black tea in a coordinate system
For position feature.Further, according to the coordinate of commodity, the distance between commodity can be calculated, since every commodity are in shelf
On placement position it is different, so the distance between each commodity difference.
It is appreciated that high-level semantic is to select the not factors shadow such as light irradiation angle in all product features of a commodity
Mark of the loud product features as the commodity can most represent the commodity using the product features that high-level semantic selects.Such as shelf
In have one bottle of black tea, body top half causes to expose due to the irradiation of light, then by high-level semantic select this bottle it is red
The product features of exposed portion can be avoided when the product features of tea to select the product features that can most represent out this bottle of black tea.
Referring to Fig. 2, step S3 is further comprised the steps of:
Step S31: coordinate system is established;
Step S32: according to the position of commodity in a coordinate system, the positional relationship between adjacent commodity is determined;
Step S33: defining phase adjacency pair in conjunction with the positional relationship between product features and commodity, and by the positional relationship between commodity
And product features are as the adjacent pairs of mark.
Coordinate system is initially set up, coordinate system is pixel coordinate system, using the lower left corner of picture as origin, to guarantee each commodity
All in the first quartile of pixel coordinate system;Then the position according to commodity in pixel coordinate system determines between adjacent commodity
Positional relationship;Finally combine positional relationship between product features and commodity to define phase adjacency pair, and by between commodity positional relationship and
Product features are as the adjacent pairs of mark, so as to match identical mark in adjacent two picture.
It is appreciated that sequence namely commodity of the positional relationship between commodity between commodity are suitable in pixel coordinate system
Sequence, as successively having black tea, green tea, laughable three commodity on shelf from left to right, then the sequence between these three commodity is commodity
Between positional relationship be labeled in conjunction with the positional relationship and the phase adjacency pair that defines of product features between commodity using semantic coding.
The product features of all pictures are more convenient to match identical mark in the first quartile of pixel coordinate system, same quadrant.
Due to the distance between each commodity difference, so each adjacent pairs of mark is different in a picture, it is similar,
The identical phase adjacency pair mark of distance in the pixel coordinate system of adjacent two picture, between product features are identical and identical product features
It is identical, in such as pixel coordinate system of adjacent two picture, select the product features of three adjacent commodity, and first figure
Three product features in piece pixel coordinate system are identical as three product features in the second picture pixel coordinate system, while the
The distance between three product features in one picture pixel coordinate system and three quotient in the second picture pixel coordinate system
The distance of product feature is identical, then uses identical mark to the phase adjacency pair that three product features in this two picture define.
Please refer to Fig. 3 A, step S33 specifically, define phase adjacency pair in conjunction with the positional relationship between product features and commodity, and
Using between commodity positional relationship and product features as the adjacent pairs of mark.It is established with the n-th picture and the (n+1)th picture
It is placed with one layer of commodity for pixel coordinate system, in shelf, is the product features selected by high-level semantic in pixel coordinate system.
It is that a phase adjacency pair has 6 in the pixel coordinate system of the n-th picture with three commodity adjacent each other in pixel coordinate system
Product features, are followed successively by A, B, B, C, B, C, corresponding pixel coordinate be followed successively by (0,0), (1,0.2), (2.1,0), (3,0.4),
(4.2,0), (4.9,0.2).Obtain 6 phase adjacency pairs by 6 product features in pixel coordinate system, i.e., (None, A, B), (A, B,
B), (B, B, C), (B, C, B), (C, B, C) and (B, C, None).In the pixel coordinate system of the (n+1)th picture, there are 6 commodity
Feature, is followed successively by C, B, C, C, D, D, corresponding pixel coordinate be followed successively by (0,0.4), (1.2,0), (1.9,0.2), (3.3,0),
(4,0.7), (5.1,0.4).Obtain 6 phase adjacency pairs by 6 adjacent product features, i.e., (None, C, B), (C, B, C), (B,
C, C), (C, C, D), (C, D, D) and (D, D, None).The left or right side that None represents the product features does not have product features.
In the pixel coordinate system of the n-th picture, the product features of 6 adjacent pairs are different, therefore to this 6 phase adjacency pairs
Using different marks;In the pixel coordinate system of (n+1)th picture, the product features of 6 adjacent pairs are different, thus to this 6
A phase adjacency pair uses different marks.Due to the phase adjacency pair (C, B, C) in the n-th picture and the phase adjacency pair in the (n+1)th picture
Product features in (C, B, C) are identical, and the distance between adjacent pairs product features are also identical, so in two pictures
Phase adjacency pair C, B, C use identical mark.It is a that (None, A, B) is marked i.e. in the n-th picture, and (A, B, B) is b, (B, B, C)
It is d for c, (B, C, B), (C, B, C) is e, and (B, C, None) is f;In the (n+1)th picture mark (None, C, B) be g, (C,
B, C) it is e, (B, C, C) is h, and (C, C, D) is i, and (C, D, D) is j, and (D, D, None) is k,.Finally according to this two pictures phase
The mark of adjacency pair, matches identical mark, i.e. e is identical mark.
Fig. 3 B is please referred to, step S4 is specifically, be sequentially mapped to same plane for the mark in all pictures, and adjacent two
All pictures are spliced into a picture, then according to the picture of splicing using overlapping mapping by identical mark in picture
The commodity on shelf are identified, that is, each commodity are identified according to the mark in splicing picture, to determine commodity on shelf.
Mark g, e, h, i, j, k adjacent pairs of in mark a, b, c, d, e, f adjacent pairs of in n-th picture and the (n+1)th picture are reflected
It is mapped to same plane, to identical phase adjacency pair e using overlapping mapping, the result of mapping is a, b, c, d, e, f, g, h, i, j, k, after
And all pictures are spliced into one, the commodity on shelf are finally gone out according to the picture recognition of splicing.
It is appreciated that having shot commodity all on shelf with the n-th picture and the (n+1)th picture, then it is spliced into
Picture contains the product features of all commodity on shelf, and the quantity of product features is equal with the quantity of commodity in shelf.
It is appreciated that the external appearance characteristic of each commodity is extracted and is stored in advance, since product features are from commodity
Extract, thus product features and commodity be it is one-to-one, due to it is adjacent pairs of be labeled as in pixel coordinate system it is adjacent each other
At least two commodity, therefore according to it is adjacent pairs of mark can get mark included in product features.Then by spliced map
The external appearance characteristic of the commodity of product features and storage in piece compares one by one, to determine which commodity there are on shelf, and can
The position that each commodity are on shelf is determined with the picture that basis is spliced into.
Further, before the commodity on shelf being shot into picture, first marked articles information, merchandise news includes commodity
External appearance characteristic, title and unit price.And then after the commodity identified, commodity can be obtained according to merchandise news on shelf
Distribution situation, proportion and commodity total price.It is appreciated that the distribution situation of commodity can be obtained according to the picture being spliced into
And proportion can obtain the total price of commodity on shelf according to the unit price of every commodity.
As a kind of deformation, pixel coordinate system can be established by origin of any position in picture, i.e. product features can be with
In any quadrant of pixel coordinate system.
The computer-readable medium that second embodiment of the invention provides is stored with computer in the computer-readable medium
Program, wherein the computer program is arranged to the method that above-mentioned identification commodity on shelf is executed when operation.
Referring to Fig. 4, the identifying system 1 that third embodiment of the invention provides comprising shooting module 11, detection module
12, coding module 13, integrate module 14;Coding module 13 includes modeling module 131, analysis module 132 and labeling module 133.
Shooting module 11 shoots the commodity on shelf at least two pictures;Detection module 12 detects the product features in picture;It builds
Mould module 131 establishes coordinate system;Position of the analysis module 132 according to commodity in a coordinate system, determines the position between adjacent commodity
Set relationship;Labeling module 133 combines the positional relationship between product features and commodity to define phase adjacency pair, and the position between commodity is closed
System and product features are as the adjacent pairs of mark;It integrates module 14 and be overlapped by identical mark in adjacent two picture and reflect
It penetrates, all pictures is spliced into a picture, to identify commodity.
In accordance with an embodiment of the present disclosure, it may be implemented as computer software journey above with reference to the process of flow chart description
Sequence.For example, embodiment of the disclosure includes a kind of computer program product comprising carry meter on a computer-readable medium
Calculation machine program, the computer program include the program code for method shown in execution flow chart.In such embodiments,
The computer program can be downloaded and installed from network by communications portion, and/or be mounted from detachable media.At this
When computer program is executed by central processing unit (CPU), the above-mentioned function of limiting in the present processes is executed.It needs to illustrate
, computer-readable medium described herein can be computer-readable signal media or computer readable storage medium
Either the two any combination.Computer readable storage medium for example may be-but not limited to-electricity, magnetic,
Optical, electromagnetic, the system of infrared ray or semiconductor, device or device, or any above combination.Computer-readable storage medium
The more specific example of matter can include but is not limited to: have the electrical connections of one or more conducting wires, portable computer diskette,
Hard disk, random access storage device (RAM), read-only memory (ROM), erasable programmable read only memory (EPROM or flash memory),
Optical fiber, portable compact disc read-only memory (CD-ROM), light storage device, magnetic memory device or above-mentioned any conjunction
Suitable combination.In this application, computer readable storage medium can be any tangible medium for including or store program, the journey
Sequence can be commanded execution system, device or device use or in connection.And in this application, it is computer-readable
Signal media may include in a base band or as carrier wave a part propagate data-signal, wherein carrying computer can
The program code of reading.The data-signal of this propagation can take various forms, including but not limited to electromagnetic signal, optical signal or
Above-mentioned any appropriate combination.Computer-readable signal media can also be any other than computer readable storage medium
Computer-readable medium, the computer-readable medium can send, propagate or transmit for by instruction execution system, device or
Person's device uses or program in connection.The program code for including on computer-readable medium can be with any appropriate
Medium transmission, including but not limited to: wireless, electric wire, optical cable, RF etc. or above-mentioned any appropriate combination.
The calculating of the operation for executing the application can be write with one or more programming languages or combinations thereof
Machine program code, described program design language include object oriented program language-such as Java, Smalltalk, C+
+, it further include conventional procedural programming language-such as " C " language or similar programming language.Program code can
Fully to execute, partly execute on the user computer on the user computer, be executed as an independent software package,
Part executes on the remote computer or executes on a remote computer or server completely on the user computer for part.
In situations involving remote computers, remote computer can pass through the network of any kind --- including local area network (LAN)
Or wide area network (WAN)-is connected to subscriber computer, or, it may be connected to outer computer (such as utilize Internet service
Provider is connected by internet).
Compared with prior art, the method for identification commodity on shelf of the invention has the advantage that
1. adjacent pairs include lap, uses phase adjacency pair as the feature of a commodity, receptive field can be expanded, avoided
Because when only using a product features and being identified, lead in same picture that the same category product features are excessively similar to be caused
Erroneous matching.
2. the influence of the product features not factors such as light irradiation angle of high-level semantic selection passes through the commodity that high-level semantic selects
Feature can more accurately identify its corresponding commodity.
3. the mark in all pictures is sequentially mapped in same plane, repeating part using overlapping mapping, then from
Distribution situation, proportion and commodity total price of the commodity on shelf can be intuitively obtained in the picture of splicing.
The foregoing is merely present pre-ferred embodiments, are not intended to limit the invention, it is all principle of the present invention it
Any modification made by interior, equivalent replacement and improvement etc. should all be comprising within protection scope of the present invention.
Claims (10)
1. a kind of method for identifying commodity on shelf, commodity on shelf for identification, it is characterised in that: this method includes following step
Suddenly,
Step S1: the commodity on shelf are shot at least two pictures;
Step S2: the product features in detection picture;
Step S3: establishing coordinate system, obtains commodity phase adjacency pair, and be labeled to phase adjacency pair;
Step S4: carrying out overlapping mapping for identical mark in adjacent two picture, all pictures be spliced into a picture,
To identify commodity.
2. the method for identification commodity on shelf as described in claim 1, it is characterised in that: step S3 is further comprised the steps of:
Step S31: coordinate system is established;
Step S32: according to the position of commodity in a coordinate system, the positional relationship between adjacent commodity is determined;
Step S33: defining phase adjacency pair in conjunction with the positional relationship between product features and commodity, and by between commodity positional relationship and quotient
Product feature is as the adjacent pairs of mark.
3. the method for identification commodity on shelf as described in claim 1, it is characterised in that: the phase adjacency pair is mutual in coordinate system
For three adjacent commodity.
4. the method for identification commodity on shelf as described in claim 1, it is characterised in that: make after the product features in detection picture
The mark for using high-level semantic to select the product features of the factors such as light irradiation angle influence as commodity.
5. the method for identification commodity on shelf as described in claim 1, it is characterised in that: the product features include the outer of commodity
See feature and position feature.
6. the method for identification commodity on shelf as claimed in claim 5, it is characterised in that: the external appearance characteristic includes the shape of commodity
Shape, color and pattern, the position feature are the coordinate of commodity in a coordinate system.
7. the method for identification commodity on shelf as described in claim 1, it is characterised in that: shooting the commodity on shelf at picture
Before, first marked articles information, merchandise news include external appearance characteristic, title and the unit price of commodity.
8. the method for identification commodity on shelf as claimed in claim 7, it is characterised in that: after the commodity identified, according to quotient
Distribution situation, proportion and commodity total price of the product information acquisition commodity on shelf.
9. a kind of computer-readable medium, it is characterised in that: it is stored with computer program in the computer-readable medium,
In, the computer program is arranged to perform claim when operation and requires identification commodity on shelf described in any one of 1-8
Method.
10. a kind of identifying system, it is characterised in that: the commodity on shelf check system includes shooting module, is used for shelf
Commodity shoot at picture;Detection module, for detecting the product features in picture;Coding module is obtained for establishing coordinate system
Commodity phase adjacency pair is obtained, and phase adjacency pair is labeled;Module is integrated, identical mark in adjacent two picture is subjected to overlapping and is reflected
It penetrates, all pictures is spliced into a picture, to identify commodity.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910055003.8A CN109934093B (en) | 2019-01-21 | 2019-01-21 | Method for identifying goods on shelf, computer readable medium and identification system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910055003.8A CN109934093B (en) | 2019-01-21 | 2019-01-21 | Method for identifying goods on shelf, computer readable medium and identification system |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109934093A true CN109934093A (en) | 2019-06-25 |
CN109934093B CN109934093B (en) | 2021-03-30 |
Family
ID=66985119
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910055003.8A Active CN109934093B (en) | 2019-01-21 | 2019-01-21 | Method for identifying goods on shelf, computer readable medium and identification system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109934093B (en) |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110489457A (en) * | 2019-07-19 | 2019-11-22 | 宜通世纪物联网研究院(广州)有限公司 | Merchandise news analysis method, system and storage medium based on image recognition |
CN111553889A (en) * | 2020-04-16 | 2020-08-18 | 上海扩博智能技术有限公司 | Method, system, equipment and storage medium for comparing commodity placement positions on goods shelf |
CN111583118A (en) * | 2020-05-13 | 2020-08-25 | 创新奇智(北京)科技有限公司 | Image splicing method and device, storage medium and electronic equipment |
CN111881894A (en) * | 2019-12-09 | 2020-11-03 | 上海扩博智能技术有限公司 | Method, system, equipment and storage medium for collecting goods selling information of container |
CN112434584A (en) * | 2020-11-16 | 2021-03-02 | 浙江大华技术股份有限公司 | Shelf commodity surplus calculation method and device, computer equipment and storage medium |
CN113674049A (en) * | 2021-06-24 | 2021-11-19 | 杭州碑石数据技术有限公司 | Commodity shelf position identification method and system based on picture search and storage medium |
CN113781371A (en) * | 2021-08-23 | 2021-12-10 | 南京掌控网络科技有限公司 | Method and equipment for removing duplication and splicing identification results of shelf pictures |
CN114511443A (en) * | 2020-10-29 | 2022-05-17 | 北京中祥英科技有限公司 | Image processing, image recognition network training and image recognition method and device |
CN116824112A (en) * | 2023-06-13 | 2023-09-29 | 广州市玄武无线科技股份有限公司 | Commodity image de-duplication identification method and device |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101442619A (en) * | 2008-12-25 | 2009-05-27 | 武汉大学 | Method for splicing non-control point image |
CN101859433A (en) * | 2009-04-10 | 2010-10-13 | 日电(中国)有限公司 | Image mosaic device and method |
US20160188994A1 (en) * | 2014-12-29 | 2016-06-30 | Intel Corporation | Method and system of feature matching for multiple images |
CN105809619A (en) * | 2015-01-19 | 2016-07-27 | 株式会社理光 | Image acquisition user interface for linear panoramic image stitching |
CN106373088A (en) * | 2016-08-25 | 2017-02-01 | 中国电子科技集团公司第十研究所 | Quick mosaic method for aviation images with high tilt rate and low overlapping rate |
CN106504192A (en) * | 2016-10-21 | 2017-03-15 | 云南电网有限责任公司电力科学研究院 | A kind of power transmission line corridor geological disaster exploration image treatment method and system |
CN108648140A (en) * | 2018-04-12 | 2018-10-12 | 上海扩博智能技术有限公司 | Image split-joint method, system, equipment and storage medium |
-
2019
- 2019-01-21 CN CN201910055003.8A patent/CN109934093B/en active Active
Patent Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101442619A (en) * | 2008-12-25 | 2009-05-27 | 武汉大学 | Method for splicing non-control point image |
CN101859433A (en) * | 2009-04-10 | 2010-10-13 | 日电(中国)有限公司 | Image mosaic device and method |
US20160188994A1 (en) * | 2014-12-29 | 2016-06-30 | Intel Corporation | Method and system of feature matching for multiple images |
CN105809619A (en) * | 2015-01-19 | 2016-07-27 | 株式会社理光 | Image acquisition user interface for linear panoramic image stitching |
US20170187953A1 (en) * | 2015-01-19 | 2017-06-29 | Ricoh Company, Ltd. | Image Acquisition User Interface for Linear Panoramic Image Stitching |
CN106373088A (en) * | 2016-08-25 | 2017-02-01 | 中国电子科技集团公司第十研究所 | Quick mosaic method for aviation images with high tilt rate and low overlapping rate |
CN106504192A (en) * | 2016-10-21 | 2017-03-15 | 云南电网有限责任公司电力科学研究院 | A kind of power transmission line corridor geological disaster exploration image treatment method and system |
CN108648140A (en) * | 2018-04-12 | 2018-10-12 | 上海扩博智能技术有限公司 | Image split-joint method, system, equipment and storage medium |
Non-Patent Citations (2)
Title |
---|
AI前线: "干货|图像算法在电商大促中的应用浅析", 《HTTPS://WWW.IYIOU.COM/P/60184.HTML》 * |
王智慧: "商品图像检测方法的研究", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
Cited By (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110489457A (en) * | 2019-07-19 | 2019-11-22 | 宜通世纪物联网研究院(广州)有限公司 | Merchandise news analysis method, system and storage medium based on image recognition |
CN111881894A (en) * | 2019-12-09 | 2020-11-03 | 上海扩博智能技术有限公司 | Method, system, equipment and storage medium for collecting goods selling information of container |
CN111553889A (en) * | 2020-04-16 | 2020-08-18 | 上海扩博智能技术有限公司 | Method, system, equipment and storage medium for comparing commodity placement positions on goods shelf |
CN111583118A (en) * | 2020-05-13 | 2020-08-25 | 创新奇智(北京)科技有限公司 | Image splicing method and device, storage medium and electronic equipment |
CN111583118B (en) * | 2020-05-13 | 2023-09-29 | 创新奇智(北京)科技有限公司 | Image stitching method and device, storage medium and electronic equipment |
CN114511443A (en) * | 2020-10-29 | 2022-05-17 | 北京中祥英科技有限公司 | Image processing, image recognition network training and image recognition method and device |
CN112434584A (en) * | 2020-11-16 | 2021-03-02 | 浙江大华技术股份有限公司 | Shelf commodity surplus calculation method and device, computer equipment and storage medium |
CN112434584B (en) * | 2020-11-16 | 2024-04-30 | 浙江大华技术股份有限公司 | Goods shelf allowance calculation method, device, computer equipment and storage medium |
CN113674049A (en) * | 2021-06-24 | 2021-11-19 | 杭州碑石数据技术有限公司 | Commodity shelf position identification method and system based on picture search and storage medium |
CN113781371A (en) * | 2021-08-23 | 2021-12-10 | 南京掌控网络科技有限公司 | Method and equipment for removing duplication and splicing identification results of shelf pictures |
CN113781371B (en) * | 2021-08-23 | 2024-01-30 | 南京掌控网络科技有限公司 | Method and equipment for de-duplication and splicing of shelf picture identification result |
CN116824112A (en) * | 2023-06-13 | 2023-09-29 | 广州市玄武无线科技股份有限公司 | Commodity image de-duplication identification method and device |
Also Published As
Publication number | Publication date |
---|---|
CN109934093B (en) | 2021-03-30 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109934093A (en) | A kind of method, computer-readable medium and identifying system identifying commodity on shelf | |
CN106920279B (en) | Three-dimensional map construction method and device | |
CN108416902B (en) | Real-time object identification method and device based on difference identification | |
CN108647553A (en) | Rapid expansion method, system, equipment and the storage medium of model training image | |
CN111814885B (en) | Method, system, device and medium for managing image frames | |
CN109753928A (en) | The recognition methods of architecture against regulations object and device | |
CN109934181A (en) | Text recognition method, device, equipment and computer-readable medium | |
CN109376731A (en) | A kind of character recognition method and device | |
CN109063653A (en) | Image processing method and device | |
CN110555879B (en) | Space positioning method, device, system and computer readable medium thereof | |
CN108389230A (en) | Refrigerator capacity automatic testing method, system, equipment and storage medium | |
CN109242060A (en) | New restocking product fast searching method, system, equipment and storage medium | |
CN106327546A (en) | Face detection algorithm test method and device | |
CN109993544A (en) | Data processing method, system, computer system and computer readable storage medium | |
CN108241561A (en) | A kind of generation method, server and the storage medium of terminal detection model | |
CN108876857A (en) | Localization method, system, equipment and the storage medium of automatic driving vehicle | |
CN109859104A (en) | A kind of video generates method, computer-readable medium and the converting system of picture | |
CN108648140A (en) | Image split-joint method, system, equipment and storage medium | |
CN109949306A (en) | Reflecting face deviation detecting method, terminal device and storage medium | |
CN112307864A (en) | Method and device for determining target object and man-machine interaction system | |
CN109961501A (en) | Method and apparatus for establishing three-dimensional stereo model | |
CN110910334B (en) | Instance segmentation method, image processing device and computer readable storage medium | |
CN111432410A (en) | Network security protection method of mobile base station of Internet of things and cloud server of Internet of things | |
CN108182457A (en) | For generating the method and apparatus of information | |
CN109901988A (en) | A kind of page elements localization method and device for automatic test |
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 |