CN110263753A - A kind of object statistical method and device - Google Patents
A kind of object statistical method and device Download PDFInfo
- Publication number
- CN110263753A CN110263753A CN201910572394.0A CN201910572394A CN110263753A CN 110263753 A CN110263753 A CN 110263753A CN 201910572394 A CN201910572394 A CN 201910572394A CN 110263753 A CN110263753 A CN 110263753A
- Authority
- CN
- China
- Prior art keywords
- counted
- image
- surrounding
- rectangular area
- converted
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- General Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Human Computer Interaction (AREA)
- Image Analysis (AREA)
Abstract
The present invention provides a kind of object statistical method and device, method includes: the image for obtaining the fixing camera shooting of monitoring area;Determine the rectangular area that each object to be counted is surrounded in image;The rectangular area for surrounding each object to be counted is converted to the preset shape for being bonded subject sizes to be counted;According to all objects to be counted in all preset shape identification and statistics images being converted to.The present invention can be accurately identified to the test object in the image for using fixing camera to shoot and quantity statistics.
Description
Technical field
The present invention relates to technical field of image processing, in particular to a kind of object statistical method and device.
Background technique
In vivo detection is the method in some authentication scenes for determining object real physiological feature.Know in face
In not applying, In vivo detection can be positioned and face by blinking, open one's mouth, shake the head, put first-class combinative movement using face key point
The technologies such as tracking verify whether user is true living body operation.Photo can be effectively resisted, changed face, mask, blocked and shield
The common attack means such as curtain reproduction.
Relative to " recognition of face ", other living animal identifications are artificial intelligence technologys under new scene and environment
Trial and exploration.Biological living identification technology is with a wide range of applications, and farm can be helped to each cultivation object,
Such as pig, ox, sheep etc., it is tracked, realizes that daily information management and whole process trace.
Summary of the invention
In view of this, can be taken the photograph to using to fix the purpose of the present invention is to provide a kind of object statistical method and device
The test object in image shot as head is accurately identified and quantity statistics.
In order to achieve the above object, the present invention provides the following technical scheme that
A kind of object statistical method, comprising:
Obtain the image of the fixing camera shooting of monitoring area;
Determine the rectangular area that each object to be counted is surrounded in image;
The rectangular area for surrounding each object to be counted is converted to the preset shape for being bonded subject sizes to be counted;
According to all objects to be counted in all preset shape identification and statistics images being converted to.
Preferably, detected using R2CNN detection model trained in advance to image, determine surrounded in image each to
The rectangular area of objects of statistics, comprising:
The horizontal rectangular frame for surrounding each object to be counted is determined using candidate region network RPN algorithm;
The characteristics of image that each horizontal rectangular frame is generated using area-of-interest pond ROI Pooling algorithm, to the figure
As feature progress regression analysis, which is adjusted to by inclined rectangular frame according to Regression Analysis Result;Described return is divided
Analysing result includes the corresponding translation of horizontal rectangular frame and rotation angle information.
Preferably, the preset shape of the fitting subject sizes to be counted is ellipse;
The rectangular area for surrounding each object to be counted is converted to the preset shape for being bonded subject sizes to be counted, is wrapped
It includes:
Elliptical central point is set by the central point for surrounding the rectangular area of the object to be counted;
Elliptical tilt angle is set by the tilt angle for surrounding the rectangular area of the object to be counted;
The width for surrounding the rectangular area of the object to be counted and height are respectively set to elliptical long axis and short axle;
It generates and meets above-mentioned central point, long axis, short axle and the ellipse of tilt angle setting, substituted and surrounded with the ellipse
The rectangular area of the object to be counted.
Preferably, according to all objects to be counted in all preset shape identification and statistics images being converted to, packet
It includes:
The each preset shape row non-maxima suppression that will be converted to, obtains the recognition result of object to be counted;
Count the quantity for determining all objects to be counted in recognition result.
A kind of object statistic device, comprising:
Acquiring unit, the image that the fixing camera for obtaining monitoring area is shot;
Determination unit, for determining the rectangular area for surrounding each object to be counted in image;
Converting unit, for the rectangular area for surrounding each object to be counted to be converted to fitting subject sizes to be counted
Preset shape;
Identification and statistic unit, in all preset shape identification and statistics images for being converted to according to converting unit
All objects to be counted.
Preferably, the determination unit, detects image using R2CNN detection model trained in advance, figure is determined
The rectangular area of each object to be counted is surrounded as in, comprising:
The horizontal rectangular frame for surrounding each object to be counted is determined using candidate region network RPN algorithm;
The characteristics of image that each horizontal rectangular frame is generated using area-of-interest pond ROI Pooling algorithm, to the figure
As feature progress regression analysis, which is adjusted to by inclined rectangular frame according to Regression Analysis Result;Described return is divided
Analysing result includes the corresponding translation of horizontal rectangular frame and rotation angle information.
Preferably, the preset shape of the fitting subject sizes to be counted is ellipse;
The rectangular area for surrounding each object to be counted is converted to fitting subject sizes to be counted by the converting unit
Preset shape, comprising:
Elliptical central point is set by the central point for surrounding the rectangular area of the object to be counted;
Elliptical tilt angle is set by the tilt angle for surrounding the rectangular area of the object to be counted;
The width for surrounding the rectangular area of the object to be counted and height are respectively set to elliptical long axis and short axle;
It generates and meets above-mentioned central point, long axis, short axle and the ellipse of tilt angle setting, substituted and surrounded with the ellipse
The rectangular area of the object to be counted.
Preferably, the identification and statistic unit, according in all preset shape identification and statistics images being converted to
All objects to be counted, comprising:
The each preset shape row non-maxima suppression that will be converted to, obtains the recognition result of object to be counted;
Count the quantity for determining all objects to be counted in recognition result.
A kind of electronic equipment, comprising: at least one processor, and be connected at least one described processor by bus
Memory;The memory is stored with the one or more computer programs that can be executed by least one described processor;Its
It is characterized in that, at least one described processor realizes above-mentioned object statistical method when executing one or more of computer programs
In step.
A kind of computer readable storage medium, the computer-readable recording medium storage one or more computer journey
Sequence, one or more of computer programs realize above-mentioned object statistical method when being executed by processor.
As can be seen from the above technical solution, in the present invention, it is to be counted right in the image shot using fixing camera to determine
After surrounding the rectangular area of each object to be counted in determining image, by the rectangular area that will surround each object to be counted
Shape conversion is carried out, it is made more to be bonded the figure of object to be counted, so as to more accurately distinguish each object to be counted
It comes, partly overlapping more than two objects to be counted can especially will be present and distinguish, thus can effectively improve
The subsequent accuracy treated when objects of statistics carries out identification and statistics.
Detailed description of the invention
Fig. 1 is object statistical method flow chart of the embodiment of the present invention;
Fig. 2 is the Object identifying result schematic diagram to be counted of the embodiment of the present invention one;
Fig. 3 is the Object identifying result schematic diagram to be counted of the embodiment of the present invention two
Fig. 4 is the structural schematic diagram of object statistic device of the embodiment of the present invention;
Fig. 5 is the structural schematic diagram of electronic equipment provided in an embodiment of the present invention.
Specific embodiment
In order to make the objectives, technical solutions, and advantages of the present invention clearer, with reference to the accompanying drawing and according to embodiment,
Technical solution of the present invention is described in detail.
In intelligently cultivation scene, the quantity of cultivation object is counted using machine vision, upper to greatest extent can reduce people
The expenditure of power resource.Technical solution provided by the invention can be used for carrying out quantity statistics to the cultivation object of intelligence cultivation scene,
To reduce cost of labor.
It is object statistical method flow chart of the embodiment of the present invention referring to Fig. 1, Fig. 1, as shown in Figure 1, this method mainly includes
Following steps:
The image that step 101, the fixing camera for obtaining monitoring area are shot.
In intelligently cultivation scene, in order to carry out quantity statistics to cultivation object, it can be disposed in each monitoring area
The fixed camera (i.e. fixing camera) in position, the image of its affiliated monitoring area is periodically shot using camera.
In the embodiment of the present invention, by obtaining image that each fixing camera is periodically shot and analyzing image, know
All cultivation objects (object i.e. to be counted) not in the affiliated monitoring scene of the fixing camera simultaneously carry out quantity statistics, can subtract
It is few to be paid by manually counting number bring human resources, reduce cost of labor.
Step 102 determines the rectangular area that each object to be counted is surrounded in image.
In practical applications, it determines the rectangular area for surrounding each object to be counted in image, image procossing can be used
Accomplished in many ways in technology.In the embodiment of the present invention, realized using R2CNN technology, it specifically, can be in advance using multiple
The training sample of object to be counted is trained to obtain R2CNN detection model, later can be used for the R2CNN detection model
In object statistic processes of the invention, specifically determined in this step using R2CNN detection model surrounded in image each to
The rectangular area of objects of statistics, it may be assumed that input an image into R2CNN detection model, R2CNN detection model carries out the image of input
Image detection surrounds the rectangular area of each object to be counted in that is, exportable image.
In the embodiment of the present invention, image detection is carried out to image using R2CNN detection model trained in advance, determines image
The middle rectangular area for surrounding each object to be counted, the specific implementation process is as follows:
The horizontal rectangular frame that each object to be counted is surrounded in image is determined using candidate region network RPN algorithm;
The characteristics of image that each horizontal rectangular frame is generated using area-of-interest pond ROI Pooling algorithm, to the figure
As feature progress regression analysis, which is adjusted to by inclined rectangular frame according to Regression Analysis Result;Described return is divided
Analysing result includes the corresponding translation of horizontal rectangular frame and rotation angle information.
Wherein, the horizontal rectangular frame that each object to be counted is surrounded in image is determined using candidate region network RPN algorithm,
The characteristics of image under different scale is mainly extracted by convolution algorithm, wherein both having included rudimentary Edge texture feature, is also wrapped
Advanced semantic feature is included, by the way that both Fusion Features get up, the complete letter for surrounding each object to be counted can be generated
Breath and the rectangle frame parallel with image boundary (referred to as horizontal rectangular frame).
It is most of existing there is no display direction to the method for In vivo detection result detected, it is only horizontally or vertically square
To testing result, and people to cultivation object carry out several when, the visual angle usually overlooked, therefore, intelligently cultivation it is this
In vivo detection in actual production scene, different from the Detection task of general goals, other than outlining cultivation object information, also
The In vivo detection towards any direction scene should be added.
Therefore, in order to sufficiently identify object information to be counted, for each horizontal rectangular frame, area-of-interest can be passed through
Pond (ROI Pooling) algorithm carries out pictorial information detection, the characteristics of image of the horizontal rectangular frame is generated, then to utilization
The characteristics of image that ROI Pooling algorithm generates carries out regression analysis, and obtained Regression Analysis Result includes horizontal rectangular frame pair
The translation and rotation angle information answered, this translation and rotation angle information show the direction tune for needing to carry out to horizontal rectangular frame
It is whole, it is the foundation that horizontal rectangular frame is adjusted to to have directive inclined rectangular frame.
The rectangular area for surrounding each object to be counted is converted to the default of fitting subject sizes to be counted by step 103
Shape.
In practical applications, can be according to the figure feature of object to be counted, object to be counted can be bonded by presetting
The shape of figure can will surround the square of the object to be counted after having determined and surrounding the rectangular area of each object to be counted
The frame conversion in shape region is bonded the shape of subject sizes to be counted so that the region of each object to be counted of encirclement outlined and
The practical region occupied of object to be counted is more bonded/close to.
In intelligently cultivation scene, cultivation object is different, and the figure for cultivating object is also not quite similar, and because fixed camera shooting
Head is usually to shoot image with depression angle, and therefore, the main position that shoots is also the torso portion for cultivating object, therefore,
In the present invention, when any cultivation object is as object to be counted, can according to cultivation object torso portion figure come into
The setting of row shape, such as it is circle, ellipse, diamond shape etc. that setting, which is bonded the shape of subject sizes to be counted,.
Pig, ox, sheep are the most common cultivation objects in intelligence cultivation scene, and the figure of torso portion is relatively ellipse
Therefore circle when object to be counted is pig, ox, any in sheep, will can be bonded the shape of subject sizes to be counted in advance
It is set as oval.
When the preset shape for being bonded subject sizes to be counted is oval, will can surround in the following ways each wait unite
The rectangular area of meter object is converted to elliptic region:
Elliptical central point is set by the central point for surrounding the rectangular area of the object to be counted;
Elliptical tilt angle is set by the tilt angle for surrounding the rectangular area of the object to be counted;
The width for surrounding the rectangular area of the object to be counted and height are respectively set to elliptical long axis and short axle;
It generates and meets above-mentioned central point, long axis, short axle and the ellipse of tilt angle setting, substituted and surrounded with the ellipse
The rectangular area of the object to be counted.
If the preset shape for being bonded subject sizes to be counted is other shapes, it is also desirable to be carried out according to true form corresponding
Conversion, but in conversion, central point, gradient require to be consistent, and the shape after conversion is no more than original rectangle
The size in region.
All objects to be counted in all preset shape identification and statistics images that step 104, basis are converted to.
In this step, in order to accurately identify all objects to be counted in image, each of will can first it be converted to pre-
If shape carries out non-maxima suppression, to obtain the recognition result of object to be counted, identification knot is then determined by statistics again
The quantity of all objects to be counted in fruit.
In practical applications, when carrying out non-maxima suppression to the region for surrounding object to be counted, for there are overlay regions
A possibility that for two objects to be counted in domain, overlapping region is bigger, one of them object to be counted is curbed is bigger, instead
It, then a possibility that curbing one of them object to be counted, is then smaller.Therefore, the application is by surrounding object to be counted
Rectangular area is converted to the preset shape for being bonded the figure of object to be counted, then carries out non-maximum suppression to preset shape again
System, it is possible to reduce the case where object to be counted of physical presence is suppressed, to effectively improve to object to be counted in image
Identification and statistics accuracy.
Fig. 2 and Fig. 3 is the cultivation object that method cultivates in scene to intelligence according to Fig. 1 respectively: pig is in different periods
The image of shooting carries out the recognition result schematic diagram obtained when identification and quantity statistics, from figures 2 and 3, it will be seen that fixation is taken the photograph
As the monitoring area of head is a swinery, all pigs in swinery can be identified using method of the invention.Here it needs
It is noted that the pig in the monitoring area shot to fixing camera is only needed to identify and quantity statistics in the present invention,
It does not need to carry out the pig except monitoring area identification and quantity statistics, therefore, in the pig except swinery (on the image of these pigs
Have the image superposition of swinery) it needs in training R2CNN detection model as this presence of duplicate sample, in this way, even if fixed camera shooting
Comprising the pig except swinery in the image of head shooting, it will not be identified and count.
Object statistical method of the embodiment of the present invention is described in detail above, the present invention also provides a kind of object systems
Counter device is described in detail below in conjunction with Fig. 4.
Referring to fig. 4, Fig. 4 is the structural schematic diagram of object statistic device of the embodiment of the present invention, as shown in figure 4, the device packet
Include acquiring unit 401, determination unit 402, converting unit 403, identification and statistic unit 404: where
Acquiring unit 401, the image that the fixing camera for obtaining monitoring area is shot;
Determination unit 402, for determining the rectangular area for surrounding each object to be counted in image;
Converting unit 403, for the rectangular area for surrounding each object to be counted to be converted to fitting subject to be counted
The preset shape of type;
Identification and statistic unit 404, all preset shape identification and statistics for being converted to according to converting unit 403
All objects to be counted in image.
In Fig. 4 shown device,
The determination unit 402 detects image using R2CNN detection model trained in advance, determines in image
Surround the rectangular area of each object to be counted, comprising:
The horizontal rectangular frame for surrounding each object to be counted is determined using candidate region network RPN algorithm;
The characteristics of image that each horizontal rectangular frame is generated using area-of-interest pond ROI Pooling algorithm, to the figure
As feature progress regression analysis, which is adjusted to by inclined rectangular frame according to Regression Analysis Result;Described return is divided
Analysing result includes the corresponding translation of horizontal rectangular frame and rotation angle information.
In Fig. 4 shown device,
The preset shape of the fitting subject sizes to be counted is ellipse;
The rectangular area for surrounding each object to be counted is converted to fitting subject to be counted by the converting unit 403
The preset shape of type, comprising:
Elliptical central point is set by the central point for surrounding the rectangular area of the object to be counted;
Elliptical tilt angle is set by the tilt angle for surrounding the rectangular area of the object to be counted;
The width for surrounding the rectangular area of the object to be counted and height are respectively set to elliptical long axis and short axle;
It generates and meets above-mentioned central point, long axis, short axle and the ellipse of tilt angle setting, substituted and surrounded with the ellipse
The rectangular area of the object to be counted.
In Fig. 4 shown device,
The identification is with statistic unit 404, all preset shape identification and statistics being converted to according to converting unit 403
All objects to be counted in image, comprising:
The each preset shape row non-maxima suppression that will be converted to, obtains the recognition result of object to be counted;
Count the quantity for determining all objects to be counted in recognition result.
The embodiment of the invention also provides a kind of electronic equipment, as shown in Figure 5, comprising: at least one processor 501, with
And the memory 502 being connected at least one described processor 501 by bus;The memory 502 is stored with can be described
One or more computer programs that at least one processor 501 executes;Wherein, at least one described processor 501 executes institute
The step in above-mentioned object statistical method shown in FIG. 1 is realized when stating one or more computer programs.
The embodiment of the invention also provides a kind of computer readable storage medium, the computer-readable recording medium storage
One or more computer programs, one or more of computer programs are realized above-mentioned shown in FIG. 1 when being executed by processor
Object statistical method.
The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the invention, all in essence of the invention
Within mind and principle, any modification, equivalent substitution, improvement and etc. done be should be included within the scope of the present invention.
Claims (10)
1. a kind of object statistical method, which is characterized in that this method comprises:
Obtain the image of the fixing camera shooting of monitoring area;
Determine the rectangular area that each object to be counted is surrounded in image;
The rectangular area for surrounding each object to be counted is converted to the preset shape for being bonded subject sizes to be counted;
According to all objects to be counted in all preset shape identification and statistics images being converted to.
2. the method according to claim 1, wherein
Image is detected using R2CNN detection model trained in advance, determines and surrounds each object to be counted in image
Rectangular area, comprising:
The horizontal rectangular frame for surrounding each object to be counted is determined using candidate region network RPN algorithm;
The characteristics of image that each horizontal rectangular frame is generated using area-of-interest pond ROI Pooling algorithm, to image spy
Sign carries out regression analysis, and the horizontal rectangular frame is adjusted to inclined rectangular frame according to Regression Analysis Result;The regression analysis knot
Fruit includes the corresponding translation of horizontal rectangular frame and rotation angle information.
3. the method according to claim 1, wherein
The preset shape of the fitting subject sizes to be counted is ellipse;
The rectangular area for surrounding each object to be counted is converted to the preset shape for being bonded subject sizes to be counted, comprising:
Elliptical central point is set by the central point for surrounding the rectangular area of the object to be counted;
Elliptical tilt angle is set by the tilt angle for surrounding the rectangular area of the object to be counted;
The width for surrounding the rectangular area of the object to be counted and height are respectively set to elliptical long axis and short axle;
Generate meet above-mentioned central point, long axis, short axle and tilt angle setting ellipse, with ellipse substitution surround be somebody's turn to do to
The rectangular area of objects of statistics.
4. the method according to claim 1, wherein
According to all objects to be counted in all preset shape identification and statistics images being converted to, comprising:
The each preset shape row non-maxima suppression that will be converted to, obtains the recognition result of object to be counted;
Count the quantity for determining all objects to be counted in recognition result.
5. a kind of object statistic device, which is characterized in that the device includes:
Acquiring unit, the image that the fixing camera for obtaining monitoring area is shot;
Determination unit, for determining the rectangular area for surrounding each object to be counted in image;
Converting unit, for the rectangular area for surrounding each object to be counted to be converted to the default of fitting subject sizes to be counted
Shape;
Identification and statistic unit, the institute in all preset shape identification and statistics images for being converted to according to converting unit
Need objects of statistics.
6. device according to claim 5, which is characterized in that
The determination unit detects image using R2CNN detection model trained in advance, determines in image and surrounds each
The rectangular area of object to be counted, comprising:
The horizontal rectangular frame for surrounding each object to be counted is determined using candidate region network RPN algorithm;
The characteristics of image that each horizontal rectangular frame is generated using area-of-interest pond ROI Pooling algorithm, to image spy
Sign carries out regression analysis, and the horizontal rectangular frame is adjusted to inclined rectangular frame according to Regression Analysis Result;The regression analysis knot
Fruit includes the corresponding translation of horizontal rectangular frame and rotation angle information.
7. device according to claim 5, which is characterized in that
The preset shape of the fitting subject sizes to be counted is ellipse;
The rectangular area for surrounding each object to be counted is converted to the default of fitting subject sizes to be counted by the converting unit
Shape, comprising:
Elliptical central point is set by the central point for surrounding the rectangular area of the object to be counted;
Elliptical tilt angle is set by the tilt angle for surrounding the rectangular area of the object to be counted;
The width for surrounding the rectangular area of the object to be counted and height are respectively set to elliptical long axis and short axle;
Generate meet above-mentioned central point, long axis, short axle and tilt angle setting ellipse, with ellipse substitution surround be somebody's turn to do to
The rectangular area of objects of statistics.
8. device according to claim 5, which is characterized in that
The identification and statistic unit, according to all to be counted in all preset shape identification and statistics images being converted to
Object, comprising:
The each preset shape row non-maxima suppression that will be converted to, obtains the recognition result of object to be counted;
Count the quantity for determining all objects to be counted in recognition result.
9. a kind of electronic equipment, comprising: at least one processor, and be connected at least one described processor by bus
Memory;The memory is stored with the one or more computer programs that can be executed by least one described processor;It is special
Sign is that at least one described processor realizes any power of claim 1-4 when executing one or more of computer programs
Method and step described in.
10. a kind of computer readable storage medium, which is characterized in that the computer-readable recording medium storage is one or more
Computer program is realized described in any one of claim 1-4 when one or more of computer programs are executed by processor
Method.
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910572394.0A CN110263753B (en) | 2019-06-28 | 2019-06-28 | Object statistical method and device |
PCT/CN2020/083513 WO2020258977A1 (en) | 2019-06-28 | 2020-04-07 | Object counting method and device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910572394.0A CN110263753B (en) | 2019-06-28 | 2019-06-28 | Object statistical method and device |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110263753A true CN110263753A (en) | 2019-09-20 |
CN110263753B CN110263753B (en) | 2020-12-22 |
Family
ID=67922783
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910572394.0A Active CN110263753B (en) | 2019-06-28 | 2019-06-28 | Object statistical method and device |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN110263753B (en) |
WO (1) | WO2020258977A1 (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2020258977A1 (en) * | 2019-06-28 | 2020-12-30 | 北京海益同展信息科技有限公司 | Object counting method and device |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115937791B (en) * | 2023-01-10 | 2023-05-16 | 华南农业大学 | Poultry counting method and device suitable for multiple cultivation modes |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108960230A (en) * | 2018-05-31 | 2018-12-07 | 中国科学院自动化研究所 | Lightweight target identification method and device based on rotation rectangle frame |
US20190012802A1 (en) * | 2017-07-06 | 2019-01-10 | Futurewei Technologies, Inc. | Recognition and reconstruction of objects with partial appearance |
CN109299688A (en) * | 2018-09-19 | 2019-02-01 | 厦门大学 | Ship Detection based on deformable fast convolution neural network |
CN109583425A (en) * | 2018-12-21 | 2019-04-05 | 西安电子科技大学 | A kind of integrated recognition methods of the remote sensing images ship based on deep learning |
CN109670501A (en) * | 2018-12-10 | 2019-04-23 | 中国科学院自动化研究所 | Object identification and crawl position detection method based on depth convolutional neural networks |
CN109685870A (en) * | 2018-11-21 | 2019-04-26 | 北京慧流科技有限公司 | Information labeling method and device, tagging equipment and storage medium |
CN109816041A (en) * | 2019-01-31 | 2019-05-28 | 南京旷云科技有限公司 | Commodity detect camera, commodity detection method and device |
Family Cites Families (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US9858496B2 (en) * | 2016-01-20 | 2018-01-02 | Microsoft Technology Licensing, Llc | Object detection and classification in images |
CN108073930A (en) * | 2017-11-17 | 2018-05-25 | 维库(厦门)信息技术有限公司 | A kind of target detection and tracking based on multiple irregular ROI |
CN109242826B (en) * | 2018-08-07 | 2022-02-22 | 高龑 | Mobile equipment end stick-shaped object root counting method and system based on target detection |
CN110263753B (en) * | 2019-06-28 | 2020-12-22 | 北京海益同展信息科技有限公司 | Object statistical method and device |
-
2019
- 2019-06-28 CN CN201910572394.0A patent/CN110263753B/en active Active
-
2020
- 2020-04-07 WO PCT/CN2020/083513 patent/WO2020258977A1/en active Application Filing
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20190012802A1 (en) * | 2017-07-06 | 2019-01-10 | Futurewei Technologies, Inc. | Recognition and reconstruction of objects with partial appearance |
CN108960230A (en) * | 2018-05-31 | 2018-12-07 | 中国科学院自动化研究所 | Lightweight target identification method and device based on rotation rectangle frame |
CN109299688A (en) * | 2018-09-19 | 2019-02-01 | 厦门大学 | Ship Detection based on deformable fast convolution neural network |
CN109685870A (en) * | 2018-11-21 | 2019-04-26 | 北京慧流科技有限公司 | Information labeling method and device, tagging equipment and storage medium |
CN109670501A (en) * | 2018-12-10 | 2019-04-23 | 中国科学院自动化研究所 | Object identification and crawl position detection method based on depth convolutional neural networks |
CN109583425A (en) * | 2018-12-21 | 2019-04-05 | 西安电子科技大学 | A kind of integrated recognition methods of the remote sensing images ship based on deep learning |
CN109816041A (en) * | 2019-01-31 | 2019-05-28 | 南京旷云科技有限公司 | Commodity detect camera, commodity detection method and device |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2020258977A1 (en) * | 2019-06-28 | 2020-12-30 | 北京海益同展信息科技有限公司 | Object counting method and device |
Also Published As
Publication number | Publication date |
---|---|
CN110263753B (en) | 2020-12-22 |
WO2020258977A1 (en) | 2020-12-30 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN105160318B (en) | Lie detecting method based on facial expression and system | |
CN105389549B (en) | Object identifying method and device based on human action feature | |
Baraldi et al. | Gesture recognition in ego-centric videos using dense trajectories and hand segmentation | |
Xiao et al. | Cow identification in free-stall barns based on an improved Mask R-CNN and an SVM | |
Theagarajan et al. | Soccer: Who has the ball? Generating visual analytics and player statistics | |
CN108549886A (en) | A kind of human face in-vivo detection method and device | |
CN108875485A (en) | A kind of base map input method, apparatus and system | |
CN106133752A (en) | Eye gaze is followed the tracks of | |
CN106372629A (en) | Living body detection method and device | |
CN108875470A (en) | The method, apparatus and computer storage medium that visitor is registered | |
CN109492534A (en) | A kind of pedestrian detection method across scene multi-pose based on Faster RCNN | |
CN108875509A (en) | Biopsy method, device and system and storage medium | |
Li et al. | Cow individual identification based on convolutional neural network | |
Błażek et al. | An unorthodox view on the problem of tracking facial expressions | |
Mussi et al. | A novel ear elements segmentation algorithm on depth map images | |
Shih | An unsupervised hair segmentation and counting system in microscopy images | |
CN110263753A (en) | A kind of object statistical method and device | |
Sankari et al. | Automatic tumor segmentation using convolutional neural networks | |
CN108875500A (en) | Pedestrian recognition methods, device, system and storage medium again | |
CN108334870A (en) | The remote monitoring system of AR device data server states | |
CN111860057A (en) | Face image blurring and living body detection method and device, storage medium and equipment | |
CN112861588B (en) | Living body detection method and device | |
CN108875501A (en) | Human body attribute recognition approach, device, system and storage medium | |
Zhang et al. | Physical activity recognition based on motion in images acquired by a wearable camera | |
CN109492585B (en) | Living body detection 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 | ||
GR01 | Patent grant | ||
CP03 | Change of name, title or address |
Address after: 601, 6 / F, building 2, No. 18, Kechuang 11th Street, Daxing District, Beijing, 100176 Patentee after: Jingdong Technology Information Technology Co.,Ltd. Address before: 601, 6 / F, building 2, No. 18, Kechuang 11th Street, Beijing Economic and Technological Development Zone, 100176 Patentee before: BEIJING HAIYI TONGZHAN INFORMATION TECHNOLOGY Co.,Ltd. |
|
CP03 | Change of name, title or address |