CN108198168A - material analyzing method and device - Google Patents
material analyzing method and device Download PDFInfo
- Publication number
- CN108198168A CN108198168A CN201711430975.8A CN201711430975A CN108198168A CN 108198168 A CN108198168 A CN 108198168A CN 201711430975 A CN201711430975 A CN 201711430975A CN 108198168 A CN108198168 A CN 108198168A
- Authority
- CN
- China
- Prior art keywords
- model
- information
- material model
- abnormal
- normal
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0012—Biomedical image inspection
- G06T7/0014—Biomedical image inspection using an image reference approach
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/12—Edge-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/90—Determination of colour characteristics
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N2021/8466—Investigation of vegetal material, e.g. leaves, plants, fruits
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30181—Earth observation
- G06T2207/30188—Vegetation; Agriculture
Abstract
The present embodiments relate to sample classification technical fields, and in particular to a kind of material analyzing method and device, method include:Obtain the image for including multiple materials taken, described image is split to obtain multiple pictures, wherein, include a material on each picture, it is respectively processed to obtain the corresponding material information of each picture for each picture, each material information is matched respectively with the normal material model to prestore and abnormal material model, to obtain the corresponding material model of each material information, described image is handled according to each material information corresponding material model.It is realized by the above method and each material in image is carried out quickly to analyze and differentiate.
Description
Technical field
The present invention relates to sample classification technical field, in particular to a kind of material analyzing method and device.
Background technology
Seed is the means of production most basic in agricultural production, is the basis of all agricultural productions.Seed it is accurate, quick
Differentiate has important directive significance to production.The domestic and international common method of seeds idenmtification mainly has fluorescent scanning identification method, changes
Learn identification method and electroresis appraisal method.These methods need stronger specialty background knowledge and longer discriminating time and can be to kinds
Son generates certain damage.Therefore, how to carry out lossless, quick discriminating to crop seeds classification is an important research side
To.
Invention content
In view of this, the purpose of the present invention is to provide a kind of material analyzing method and devices, and material is carried out with realizing
It is lossless rapidly to differentiate.
Present pre-ferred embodiments provide a kind of material analyzing method, including:
Obtain the image for including multiple materials taken;
Described image is split to obtain multiple pictures, wherein, include a material on each picture;
It is respectively processed to obtain the corresponding material information of each picture for each picture;
Each material information is matched respectively with the normal material model to prestore and abnormal material model, to obtain
The corresponding material model of each material information;
Described image is handled according to the material information corresponding material model.
Optionally, in above-mentioned material analyzing method, the step of being split described image to obtain multiple pictures, wraps
It includes:
For each material in described image, the positioning of different angle at least twice is carried out based on YOLO2 algorithms, with root
The frame of each material is obtained according to the positioning of the different angle at least twice;
Described image is split according to the frame of each material to obtain multiple pictures.
Optionally, in above-mentioned material analyzing method, by each material information and the normal material model that prestores and different
Normal material model is matched respectively, to include the step of obtaining each material information corresponding material model:
It is carried out for each material information according to VGG16 structures, the normal material model and abnormal material model
Classify to obtain the probability of the corresponding different material model of each material information;
The corresponding material model of the material information is obtained according to the probability of the corresponding different material model of the material information,
Wherein, the corresponding material model of the material information is normal material model or abnormal material model.
Optionally, in above-mentioned material analyzing method, the normal material model includes normal shape model and normal face
Color model, the exception material model includes abnormal shape model and abnormal color model, described for each picture point
It is not handled to include the step of obtaining each picture corresponding material information:
SHAPE DETECTION and color identification are carried out respectively for the material in each picture to obtain each material
Shape information and colouring information;
The step of each material information is matched respectively with the normal material model to prestore and abnormal material model
Including:
For each material, by the corresponding colouring information of the material and the normal color model and abnormal color model into
Row matching, the corresponding shape information of the material is matched respectively with the normal shape model and abnormal shape model.
Optionally, in above-mentioned material analyzing method, the exception material model includes worm-eaten material model, scab material
Model, broken kernel material model, sprouted kernel material model and the material model that mildews by each material information and prestore
The step of normal material model and abnormal material model are matched respectively includes:
By each material information and the normal material model, worm-eaten material model, scab material model, broken kernel object
Model, sprouted kernel material model and the material model that mildews are expected, to obtain and each highest object of material information matching degree
Expect model.
Optionally, in above-mentioned material analyzing method, according to the corresponding material model of each material information to the figure
Include as the step of being handled:
When the corresponding material model of the material information is abnormal material model, by described image with the material information
Corresponding material is marked;
The method further includes:
The material quantity included according to the material quantity of label and described image obtains the probability of abnormal material and output.
The present invention also provides a kind of material analyzing device, described device includes:
Image collection module, for obtaining the image for including multiple materials taken;
Divide module, for being split described image to obtain multiple pictures, wherein, include on each picture
One material;
Data obtaining module is respectively processed to obtain the corresponding object of each picture for being directed to each picture
Expect information;
Matching module, for by each material information and the normal material model that prestores and exception material model respectively into
Row matching, to obtain the corresponding material model of each material information;
Processing module, for being handled according to the corresponding material model of the material information described image.
Optionally, in above-mentioned material analyzing device, the segmentation module includes:
Submodule is positioned, for being directed to each material in described image, is carried out based on YOLO2 algorithms different at least twice
The positioning of angle, with the positioning of different angle obtains the frame of each material at least twice according to;
Divide submodule, for being split to obtain multiple pictures to described image according to the frame of each material.
Optionally, in above-mentioned material analyzing device, the matching module includes:
Computational submodule, for being directed to each material information according to VGG16 structures, the normal material model and different
Normal material model is classified with the probability for obtaining the corresponding different material model of each material information;
Submodule is obtained, for obtaining the material information according to the probability of the corresponding different material model of the material information
Corresponding material model, wherein, the corresponding material model of the material information is normal material model or abnormal material model.
Optionally, in above-mentioned material analyzing device, the processing module is additionally operable to when the corresponding object of the material information
When expecting model for abnormal material model, material corresponding with the material information in described image is marked;
The material analyzing device further includes statistical module, includes for the material quantity according to label and described image
Material quantity obtain probability and the output of abnormal material.
A kind of material analyzing method and device provided by the invention, by the image including multiple materials to getting into
Row segmentation so that obtained each picture after segmentation includes a material, and is analyzed each picture and with prestoring
Normal material model and exception material model are matched the corresponding material model of material information to obtain each picture respectively, with
Image is handled, and then realize the analysis to material each in image according to each material information corresponding material model
And discriminating.
For the above objects, features and advantages of the present invention is enable to be clearer and more comprehensible, preferred embodiment cited below particularly, and coordinate
Appended attached drawing, is described in detail below.
Description of the drawings
It in order to illustrate the technical solution of the embodiments of the present invention more clearly, below will be to needed in the embodiment attached
Figure is briefly described, it should be understood that the following drawings illustrates only certain embodiments of the present invention, therefore is not construed as pair
The restriction of range, for those of ordinary skill in the art, without creative efforts, can also be according to this
A little attached drawings obtain other relevant attached drawings.
Fig. 1 is the connection block diagram of a kind of electronic equipment provided in an embodiment of the present invention.
Fig. 2 is a kind of flow diagram of material analyzing method provided in an embodiment of the present invention.
Fig. 3 is the step schematic diagram of step S120 in Fig. 2.
Fig. 4 is the step schematic diagram of step S140 in Fig. 2.
Fig. 5 is a kind of connection block diagram of material analyzing device provided in an embodiment of the present invention.
Fig. 6 is the connection block diagram of segmentation module provided in an embodiment of the present invention.
Fig. 7 is the connection block diagram of matching module provided in an embodiment of the present invention.
Icon:10- electronic equipments;12- memories;14- processors;100- material analyzing devices;110- image acquisition moulds
Block;120- divides module;122- positions submodule;124- divides submodule;130- data obtaining modules;140- matching modules;
142- computational submodules;144- obtains submodule;150- processing modules;160- statistical modules.
Specific embodiment
Below in conjunction with attached drawing in the embodiment of the present invention, the technical solution in the embodiment of the present invention is carried out clear, complete
Ground describes, it is clear that described embodiment is only part of the embodiment of the present invention, instead of all the embodiments.Usually exist
The component of the embodiment of the present invention described and illustrated in attached drawing can be configured to arrange and design with a variety of different herein.
Therefore, below the detailed description of the embodiment of the present invention to providing in the accompanying drawings be not intended to limit it is claimed
The scope of the present invention, but be merely representative of the present invention selected embodiment.Based on the embodiment of the present invention, people in the art
Member's all other embodiments obtained under the premise of creative work is not made, shall fall within the protection scope of the present invention.
At present, the conventional method for crops being carried out with pest diagnosis is mainly to be estimated according to artificial, however this peasant
The judgement rule of thumb made does not ensure that completely correct, and there are the shortcomings such as subjectivity is strong, heavy workload, efficiency are low;Separately
On the one hand, due to there is no professional person to arrive field diagnostic in time, the state of an illness may be made to be delayed or aggravate.
As shown in Figure 1, be present pre-ferred embodiments provide electronic equipment 10 block diagram, the electronic equipment
10 include memory 12, processor 14 and display 16.
The memory 12, processor 14 and display 16 are directly or indirectly electrically connected between each other, to realize number
According to transmission or interaction.For example, these elements can be realized electrically between each other by one or more communication bus or signal wire
Connection.The software function that the memory 12 is stored in the form of software or firmware (Firmware) is stored in memory 12
Module, the processor 14 is stored in software program and module in memory 12 by operation, in the embodiment of the present invention
Material analyzing device 100, so as to perform various functions application and data processing, that is, realize the material in the embodiment of the present invention
Analysis method.
In the present embodiment, the electronic equipment 10 refers to the hardware device with data-handling capacity.The electronic equipment
10 can be mobile phone, computer, tablet computer or the arbitrary equipment with data-handling capacity and display capabilities, not make herein specific
It limits.
The memory 12 may be, but not limited to, random access memory (Random Access Memory, RAM),
Read-only memory (Read Only Memory, ROM), programmable read only memory (Programmable Read-Only
Memory, PROM), erasable read-only memory (Erasable Programmable Read-Only Memory, EPROM),
Electricallyerasable ROM (EEROM) (Electric Erasable Programmable Read-Only Memory, EEPROM) etc..
The memory 12 is for storing program, for example, the function corresponding to material analyzing device 100 provided by the invention
Module can perform the function module to realize that the type to material is analyzed by the processor 14.
It is appreciated that structure shown in FIG. 1 is only to illustrate, electronic equipment 10 may also include it is more than shown in Fig. 1 or
Less component or with the configuration different from shown in Fig. 1.Each component shown in Fig. 1 may be used hardware, software or its
Combination is realized.
Referring to Fig. 2, the present invention provides a kind of material analyzing method, the material analyzing method is set applied to the electronics
Standby 10, the method is applied to perform step S110- steps five steps of S150 during the electronic equipment 10.
Step S110:Obtain the image for including multiple materials taken.
Wherein, the image including multiple materials can be taken by the electronic equipment 10 or by with electricity
What the image collecting device that sub- equipment 10 connects took, it is not specifically limited herein.Before being shot, the multiple object
A plane is placed during material by the way of tiling, so that in the image including multiple materials taken, between each material
It is separated from each other.The material can be vegetable seeds or industrial part etc., be not specifically limited herein.Optionally, exist
In the present embodiment, the material is seed, and the seed can be but not limited to soybean, corn, wheat or paddy etc..
Step S120:Described image is split to obtain multiple pictures, wherein, include one on each picture
Material.
Wherein, can be that each material in described image is positioned respectively to the mode that described image is split,
And the picture after positioning is divided using circular dividing or square;Can also be that described image is detected to be each got
The profile information of each material into described image, is split described image according to the profile information.Do not make herein specific
It limits, is chosen according to actual demand.
Incorporated by reference to Fig. 3, optionally, in the present embodiment, the step of being split described image to obtain multiple pictures
Including:
Step S122:For each material in described image, different angle at least twice is carried out based on YOLO2 algorithms
Positioning, with the positioning of different angle obtains the frame of each material at least twice according to.
Step S124:Described image is split according to the frame of each material to obtain multiple pictures.
Wherein, the number at least twice can be but not limited to two, three times or four times.With time at least twice
Number is illustrates for twice, and for each material in described image, different angle twice is carried out based on YOLO2 algorithms
Positioning obtains two division frames to realize for each material, and obtains two parts for dividing frame overlapping and obtain each material
Corresponding frame.The angular range of two division frames that the positioning of different angle twice obtains can be the root between 0-90 degree
It is configured, is not specifically limited herein according to the shape of the material.To make the better of the picture after division, at this
In embodiment, the angle of described two frames is 45 degree.
By above-mentioned setting effectively to avoid when handling image, it is possible to prevente effectively from each object in described image
Material influences the shape of the material in the picture because being divided after obtaining picture to image when spaced too close together or presence partly overlaps
Shape, and then influence the analysis result to the material in picture.
Step S130:It is respectively processed to obtain the corresponding material information of each picture for each picture.
Wherein, the corresponding material information of the picture can include the corresponding colouring information of material in the picture and institute
State the corresponding shape information of picture.Wherein, which can be a kind of or a variety of.
In the present embodiment, it is described to be respectively processed to obtain the corresponding object of each picture for each picture
The step of expecting information includes:SHAPE DETECTION and color identification are carried out respectively for each picture to obtain each picture
Shape information and colouring information.
Step S140:Each material information and the normal material model that prestores and exception material model are carried out respectively
Match, to obtain the corresponding material model of each material information.
Wherein, the normal material model includes normal shape model and normal color model, the exception material model
Including abnormal shape model and abnormal color model.By each material information and the normal material model to prestore and abnormal material
Model carries out matched mode respectively:For each material, by the corresponding colouring information of the material and the normal face
Color model and abnormal color model are matched, by the corresponding shape information of the material and the normal shape model and abnormal shape
Shape model is matched respectively.Specifically, when the corresponding colouring information of the picture and the matching degree of the normal color model are big
In the matching degree with abnormal color, and the corresponding shape information of material in the picture and the matching degree of normal shape model are more than
During with the matching degree of normal shape model, then the corresponding material model of the material in the picture is normal material model, otherwise for
Abnormal material model.By above-mentioned setting to realize the analysis to material each in described image.
Optionally, in the present embodiment, the abnormal material model can include worm-eaten material model, scab material mould
Type, broken kernel material model, sprouted kernel material model and the material model that mildews by each material information and prestore just
The step of normal material model and abnormal material model are matched respectively includes:By each material information and the normal material
Model, worm-eaten material model, scab material model, broken kernel material model, sprouted kernel material model and the material mould that mildews
Type, to obtain and each highest material model of material information matching degree.
By above-mentioned setting to realize the further analysis to different abnormal material models, to obtain the shape of the material
State, the i.e. material are normal, worm-eaten, scab, crush, sprout or mildew state.And then effective guarantee distinguishes the accuracy of result.
Incorporated by reference to Fig. 4, to further improve the efficiency and accuracy analyzed material each in image, optionally,
In the present embodiment, each material information is matched respectively with the normal material model to prestore and abnormal material model, with
The step of obtaining each material information corresponding material model also includes:
Step S142:For each material information according to VGG16 structures, the normal material model and abnormal material
Model is classified with the probability for obtaining the corresponding different material model of each material information.
Step S144:It is corresponding that the material information is obtained according to the probability of the corresponding different material model of the material information
Material model, wherein, the corresponding material model of the material information is normal material model or abnormal material model.
Optionally, in the present embodiment, the abnormal material model can include worm-eaten material model, scab material mould
Type, broken kernel material model, sprouted kernel material model and the material model that mildews.For each material information according to
VGG16 structures, the normal material model classify to obtain the corresponding difference of each material information with abnormal material model
The step of probability of material model, includes:For each material information according to VGG16 structures, normal material model, worm-eaten
Material model, scab material model, broken kernel material model, sprouted kernel material model and the material model that mildews pass through above-mentioned
Setting is to realize the classification to different abnormal material models, to obtain the state of the material, i.e., the material be normal, worm-eaten,
Scab, crush, sprout or mildew state.And then the accuracy of result that effective guarantee analyzes material.In addition, by adopting
In aforementioned manners with effectively increase to material each in described image carry out analysis component efficiency and analysis accuracy.
Step S150:Described image is handled according to each material information corresponding material model.
Wherein, can be according to the mode that the corresponding material model of the material information handles described image, when
When the corresponding material model of the material information is abnormal material model, by material corresponding with the material information in described image
It is marked.Wherein, when the abnormality difference of the corresponding material of different material informations, to corresponding to different material into rower
The label information of generation of clocking can be identical or different, and the mode being marked can be carried out with word
It identifies or symbolization is identified, be not specifically limited herein.By above-mentioned setting so that user can be intuitive
Check the normal material in image and abnormal material.
For further facilitate user understand to the material analyzed as a result, optional, it is in the present embodiment, described
Method further includes:The material quantity included according to the material quantity of label and described image obtains the probability of abnormal material and defeated
Go out.
Wherein, when the abnormal material includes worm-eaten, scab, crushes, sprouts or mildews, according to labeled as worm-eaten, scab,
The quantity for the material for crushing, sprouting or mildewing and labeled quantity obtain different unusual conditions in the case where accounting for total material unusual condition
Probability and the material in the quantity and described image labeled as worm-eaten, the material for scab, crushing, sprouting or mildewing
The material quantity that quantity obtains worm-eaten, scab, crushes, sprouts or mildews accounts for the ratio of total material quantity in image.
Incorporated by reference to Fig. 5, the present invention also provides a kind of material analyzing device 100, including image collection module 110, segmentation mould
Block 120, data obtaining module 130, matching module 140 and processing module 150
The image for including multiple materials that described image acquisition module 110 takes for acquisition.Specifically, described image
Acquisition module 110 can be used for performing step S110 shown in Fig. 2, and specific operating method can refer to retouching in detail for step S110
It states.
The segmentation module 120 is used to be split described image to obtain multiple pictures, wherein, each picture
It is upper to include a material.Specifically, the segmentation module 120 can be used for performing sub-step S120 shown in Fig. 2, specific to grasp
It can refer to the detailed description of step S120 as method.
Incorporated by reference to Fig. 6, optionally, in the present embodiment, the segmentation module 120 includes positioning submodule 122 and segmentation
Submodule 124.
The positioning submodule 122 is used for for each material in described image, and at least two are carried out based on YOLO2 algorithms
The positioning of secondary different angle, with the positioning of different angle obtains the frame of each material at least twice according to.Specifically
Ground, the positioning submodule 122 can be used for performing the sub-step S122 shown in Fig. 3, and specific operating method can refer to step
The detailed description of S124.
The segmentation submodule 124 is multiple to obtain for being split according to the frame of each material to described image
Picture.Specifically, the segmentation submodule 124 can be used for performing the sub-step S124 shown in Fig. 3, and specific operating method can
With reference to the detailed description of step S124.
Described information acquisition module 130 is used to be respectively processed to obtain each picture pair for each picture
The material information answered.Specifically, described information acquisition module 130 can be used for performing sub-step S130 shown in Fig. 2, specifically
Operating method can refer to the detailed description of step S130.
The matching module 140 is used for each material information and the normal material model to prestore and abnormal material model
It is matched respectively, to obtain the corresponding material model of each material information.Specifically, the matching module 140 can be used for
Sub-step S140 shown in Fig. 2 is performed, specific operating method can refer to the detailed description of step S140.
Incorporated by reference to Fig. 7, optionally, in the present embodiment, the matching module 140 includes computational submodule 142 and obtains
Submodule 144.
The computational submodule 142 is used for for each material information according to VGG16 structures, the normal material mould
Type is classified with abnormal material model with the probability for obtaining the corresponding different material model of each material information.Specifically, institute
It states computational submodule 142 to can be used for performing the sub-step S142 shown in Fig. 4, specific operating method can refer to step S142's
Detailed description.
The acquisition submodule 144 is used to obtain the object according to the probability of the corresponding different material model of the material information
Expect the corresponding material model of information.Specifically, the submodule 144 that obtains can be used for performing the sub-step S144 shown in Fig. 4,
Specific operating method can refer to the detailed description of step S144.Wherein, the corresponding material model of the material information is normal object
Expect model or abnormal material model
The processing module 150 is used to handle described image according to the corresponding material model of the material information.
Specifically, the processing module 150 can be used for performing sub-step S150 shown in Fig. 2, and specific operating method can refer to step
The detailed description of rapid S150.
Optionally, the processing module 150 is additionally operable to when the corresponding material model of the material information in the present embodiment
During for abnormal material model, material corresponding with the material information in described image is marked, the material analyzing device
100 further include statistical module 160.
The material quantity that the statistical module 160 is used to be included according to the material quantity and described image of label obtains different
The probability of normal material and output.Specifically, the specific implementation process of the statistical module 160 please refers to above-mentioned material analyzing side
The description of method, does not repeat one by one herein.
To sum up, a kind of material analyzing method and device provided by the invention, method include to acquisition take include it is more
The image of a material is split to obtain multiple pictures, wherein, include a material on each picture, for each institute
Picture is stated to be respectively processed to obtain the corresponding material information of each picture, and by each material information and prestore just
Normal material model and abnormal material model are matched respectively, to obtain the corresponding material model of each material information, according to
The corresponding material model of each material information handles described image, to realize to material analyzing each in image and mirror
Not.
In embodiment provided herein, it should be understood that disclosed device and method, it can also be by other
Mode realize.The apparatus embodiments described above are merely exemplary, for example, the flow chart and block diagram in attached drawing are shown
Architectural framework in the cards, function and the behaviour of devices in accordance with embodiments of the present invention, method and computer program product
Make.
It should be noted that herein, term " comprising ", "comprising" or its any other variant are intended to non-row
His property includes, so that process, method, article or equipment including a series of elements not only include those elements, and
And it further includes other elements that are not explicitly listed or further includes intrinsic for this process, method, article or equipment institute
Element.In the absence of more restrictions, the element limited by sentence "including a ...", it is not excluded that including institute
State in process, method, article or the equipment of element that also there are other identical elements.
The foregoing is only a preferred embodiment of the present invention, is not intended to restrict the invention, for the skill of this field
For art personnel, the invention may be variously modified and varied.All within the spirits and principles of the present invention, that is made any repaiies
Change, equivalent replacement, improvement etc., should all be included in the protection scope of the present invention.It should be noted that:Similar label and letter exists
Similar terms are represented in following attached drawing, therefore, once being defined in a certain Xiang Yi attached drawing, are then not required in subsequent attached drawing
It is further defined and is explained.
The above description is merely a specific embodiment, but protection scope of the present invention is not limited thereto, any
Those familiar with the art in the technical scope disclosed by the present invention, can readily occur in change or replacement, should all contain
Lid is within protection scope of the present invention.Therefore, protection scope of the present invention should be based on the protection scope of the described claims.
Claims (10)
- A kind of 1. material analyzing method, which is characterized in that the method includes:Obtain the image for including multiple materials taken;Described image is split to obtain multiple pictures, wherein, include a material on each picture;It is respectively processed to obtain the corresponding material information of each picture for each picture;Each material information is matched respectively with the normal material model to prestore and abnormal material model, to obtain each institute State the corresponding material model of material information;Described image is handled according to each material information corresponding material model.
- 2. material analyzing method according to claim 1, which is characterized in that be split to described image multiple to obtain The step of picture, includes:For each material in described image, the positioning of different angle at least twice is carried out based on YOLO2 algorithms, with according to institute The positioning for stating different angle at least twice obtains the frame of each material;Described image is split according to the frame of each material to obtain multiple pictures.
- 3. material analyzing method according to claim 1, which is characterized in that each material information is normal with prestoring Material model and abnormal material model are matched respectively, to be wrapped the step of obtaining each material information corresponding material model It includes:Classified for each material information according to VGG16 structures, the normal material model and abnormal material model To obtain the probability of the corresponding different material model of each material information;The corresponding material model of the material information is obtained according to the probability of the corresponding different material model of the material information, In, the corresponding material model of the material information is normal material model or abnormal material model.
- 4. material analyzing method according to claim 1, which is characterized in that the normal material model includes normal shape Model and normal color model, the exception material model includes abnormal shape model and abnormal color model, described for every A picture is respectively processed to include the step of obtaining each picture corresponding material information:SHAPE DETECTION and color identification are carried out respectively for the material in each picture to obtain the shape of each material Information and colouring information;The step of each material information is matched respectively with the normal material model to prestore and abnormal material model includes:For each material, by the corresponding colouring information of the material and the progress of the normal color model and abnormal color model Match, the corresponding shape information of the material is matched respectively with the normal shape model and abnormal shape model.
- 5. material analyzing method according to claim 1, which is characterized in that the exception material model includes worm-eaten material Model, scab material model, broken kernel material model, sprouted kernel material model and the material model that mildews, by each object The step of material information is matched respectively with the normal material model to prestore and abnormal material model includes:By each material information and the normal material model, worm-eaten material model, scab material model, broken kernel material mould Type, sprouted kernel material model and the material model that mildews, to obtain and each highest material mould of material information matching degree Type.
- 6. material analyzing method according to claim 1, which is characterized in that according to the corresponding material of each material information The step of model handles described image includes:It, will be corresponding with the material information in described image when the corresponding material model of the material information is abnormal material model Material be marked;The method further includes:The material quantity included according to the material quantity of label and described image obtains the probability of abnormal material and output.
- 7. a kind of material analyzing device, which is characterized in that described device includes:Image collection module, for obtaining the image for including multiple materials taken;Divide module, for being split described image to obtain multiple pictures, wherein, include one on each picture Material;Data obtaining module is respectively processed to obtain the corresponding material letter of each picture for being directed to each picture Breath;A matching module, for each material information and the normal material model that prestores and exception material model to be carried out respectively Match, to obtain the corresponding material model of each material information;Processing module, for being handled according to the corresponding material model of the material information described image.
- 8. material analyzing device according to claim 7, which is characterized in that the segmentation module includes:Submodule is positioned, for being directed to each material in described image, different angle at least twice is carried out based on YOLO2 algorithms Positioning, the frame of each material is obtained with the positioning of the different angle at least twice according to;Divide submodule, for being split to obtain multiple pictures to described image according to the frame of each material.
- 9. material analyzing device according to claim 8, which is characterized in that the matching module includes:Computational submodule, for being directed to each material information according to VGG16 structures, the normal material model and anomalies Material model is classified with the probability for obtaining the corresponding different material model of each material information;Submodule is obtained, is corresponded to for obtaining the material information according to the probability of the corresponding different material model of the material information Material model, wherein, the corresponding material model of the material information is normal material model or abnormal material model.
- 10. material analyzing device according to claim 7, which is characterized in that the processing module is additionally operable to when the object When to expect the corresponding material model of information be abnormal material model, by material corresponding with the material information in described image into rower Note;The material analyzing device further includes statistical module, the object included for the material quantity according to label and described image Material quantity obtains probability and the output of abnormal material.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711430975.8A CN108198168A (en) | 2017-12-26 | 2017-12-26 | material analyzing method and device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711430975.8A CN108198168A (en) | 2017-12-26 | 2017-12-26 | material analyzing method and device |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108198168A true CN108198168A (en) | 2018-06-22 |
Family
ID=62584103
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201711430975.8A Pending CN108198168A (en) | 2017-12-26 | 2017-12-26 | material analyzing method and device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108198168A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109030509A (en) * | 2018-07-02 | 2018-12-18 | 京东方科技集团股份有限公司 | A kind of detection device, color film prosthetic appliance and detection method |
Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101234381A (en) * | 2008-03-07 | 2008-08-06 | 天津市华核科技有限公司 | Granular material sorting classifying method based on visual sense recognition |
CN102059221A (en) * | 2010-10-15 | 2011-05-18 | 合肥泰禾光电科技有限公司 | Color material separation device and method |
CN104463265A (en) * | 2013-09-13 | 2015-03-25 | 佳益私人有限公司 | Material handling system and method |
CN105117734A (en) * | 2015-07-28 | 2015-12-02 | 江南大学 | Corn seed hyper-spectral image classification identification method based on model on-line updating |
CN105628708A (en) * | 2016-01-27 | 2016-06-01 | 塔里木大学 | Quick nondestructive testing method for multi-parameter quality of south Xinjiang red dates |
CN105954281A (en) * | 2016-04-21 | 2016-09-21 | 南京农业大学 | Method for non-destructive identification of paddy moldy fungal colony |
CN106295705A (en) * | 2016-08-17 | 2017-01-04 | 汕头市三三智能科技有限公司 | A kind of many colors material screening number system under movement background |
CN106975617A (en) * | 2017-04-12 | 2017-07-25 | 北京理工大学 | A kind of Classification of materials method based on color selector |
CN107256262A (en) * | 2017-06-13 | 2017-10-17 | 西安电子科技大学 | A kind of image search method based on object detection |
CN107330437A (en) * | 2017-07-03 | 2017-11-07 | 贵州大学 | Feature extracting method based on the real-time detection model of convolutional neural networks target |
-
2017
- 2017-12-26 CN CN201711430975.8A patent/CN108198168A/en active Pending
Patent Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101234381A (en) * | 2008-03-07 | 2008-08-06 | 天津市华核科技有限公司 | Granular material sorting classifying method based on visual sense recognition |
CN102059221A (en) * | 2010-10-15 | 2011-05-18 | 合肥泰禾光电科技有限公司 | Color material separation device and method |
CN104463265A (en) * | 2013-09-13 | 2015-03-25 | 佳益私人有限公司 | Material handling system and method |
CN105117734A (en) * | 2015-07-28 | 2015-12-02 | 江南大学 | Corn seed hyper-spectral image classification identification method based on model on-line updating |
CN105628708A (en) * | 2016-01-27 | 2016-06-01 | 塔里木大学 | Quick nondestructive testing method for multi-parameter quality of south Xinjiang red dates |
CN105954281A (en) * | 2016-04-21 | 2016-09-21 | 南京农业大学 | Method for non-destructive identification of paddy moldy fungal colony |
CN106295705A (en) * | 2016-08-17 | 2017-01-04 | 汕头市三三智能科技有限公司 | A kind of many colors material screening number system under movement background |
CN106975617A (en) * | 2017-04-12 | 2017-07-25 | 北京理工大学 | A kind of Classification of materials method based on color selector |
CN107256262A (en) * | 2017-06-13 | 2017-10-17 | 西安电子科技大学 | A kind of image search method based on object detection |
CN107330437A (en) * | 2017-07-03 | 2017-11-07 | 贵州大学 | Feature extracting method based on the real-time detection model of convolutional neural networks target |
Non-Patent Citations (2)
Title |
---|
JOSEPH REDMON ET AL: "YOLO9000: Better, Faster, Stronger", 《2017 IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR)》 * |
孙群等: "无损检测技术在种子质量检验上的应用研究进展", 《中国农业大学学报》 * |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109030509A (en) * | 2018-07-02 | 2018-12-18 | 京东方科技集团股份有限公司 | A kind of detection device, color film prosthetic appliance and detection method |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Pantazi et al. | Active learning system for weed species recognition based on hyperspectral sensing | |
Aquino et al. | A new methodology for estimating the grapevine-berry number per cluster using image analysis | |
Ali et al. | Symptom based automated detection of citrus diseases using color histogram and textural descriptors | |
Bauriegel et al. | Hyperspectral and chlorophyll fluorescence imaging for early detection of plant diseases, with special reference to Fusarium spec. infections on wheat | |
Costa et al. | Shape analysis of agricultural products: a review of recent research advances and potential application to computer vision | |
Bai et al. | Rice heading stage automatic observation by multi-classifier cascade based rice spike detection method | |
Bauriegel et al. | Hyperspectral and chlorophyll fluorescence imaging to analyse the impact of Fusarium culmorum on the photosynthetic integrity of infected wheat ears | |
Yang et al. | Hyperspectral band selection for detecting different blueberry fruit maturity stages | |
EP2854053A1 (en) | Defect prediction method and device | |
Mesa et al. | Multi-input deep learning model with RGB and hyperspectral imaging for banana grading | |
CN107222682A (en) | Crop growth state testing method and device | |
CN106778782A (en) | Identifying pest and prevention and controls and system | |
Przybył et al. | The recognition of potato varieties using of neural image analysis method | |
Zhang et al. | High-throughput corn ear screening method based on two-pathway convolutional neural network | |
Shaikh et al. | Citrus leaf unhealthy region detection by using image processing technique | |
Ye et al. | An image-based approach for automatic detecting tasseling stage of maize using spatio-temporal saliency | |
Scalisi et al. | Maturity prediction in yellow peach (Prunus persica L.) cultivars using a fluorescence spectrometer | |
Daszykowski et al. | Robust SIMCA-bounding influence of outliers | |
CN116452316A (en) | Post-loan risk monitoring method and device for agricultural loan and electronic equipment | |
Segal et al. | Identifying complex sources in large astronomical data sets using a coarse-grained complexity measure | |
US8229879B2 (en) | System and method for auto-adaptive network | |
Sangeetha et al. | A novel exploration of plant disease and pest detection using machine learning and deep learning algorithms | |
CN108198168A (en) | material analyzing method and device | |
EP3301633B1 (en) | Remote chemical assay classification | |
Pujari et al. | Detection and classification of fungal disease with radon transform and support vector machine affected on cereals |
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180622 |
|
RJ01 | Rejection of invention patent application after publication |