CN203177944U - Particle counting apparatus - Google Patents
Particle counting apparatus Download PDFInfo
- Publication number
- CN203177944U CN203177944U CN 201320172524 CN201320172524U CN203177944U CN 203177944 U CN203177944 U CN 203177944U CN 201320172524 CN201320172524 CN 201320172524 CN 201320172524 U CN201320172524 U CN 201320172524U CN 203177944 U CN203177944 U CN 203177944U
- Authority
- CN
- China
- Prior art keywords
- particle
- counting
- particle counting
- image
- transparent pallet
- 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.)
- Expired - Lifetime
Links
Images
Landscapes
- Image Processing (AREA)
Abstract
The utility model discloses a particle counting apparatus, comprising a transparent pallet and material illumination light sources arranged at the periphery of the transparent pallet. Camera equipment is arranged over the transparent pallet, the camera equipment is connected with a computer through a data line, the equipment driver of the camera equipment is called through an interface, a keyboard and a mouse are employed for controlling shooting, and the material illumination light sources are white LED light sources with angles and light intensity adjustable. The particle counting apparatus is conveniently applied for particle counting instead of manual counting, work efficiency is greatly raised, manual dependence degree is low in the whole detection process, good repeatability and high efficiency are provided, simultaneously, the requirement for image quality is not high, and the particle counting apparatus is widely applied for counting field.
Description
Technical field
The utility model belongs to areas of information technology, relates to a kind of particle counting device.
Background technology
So-called mass of 1000 kernel is the weight that restrains 1,000 grain (seed) of expression, and it is an index that embodies grain size and full degree, is the content of check grain quality and crop species test, also is the important evaluation index in the wine brewing process.Mensuration for the mass of 1000 kernel index mainly realizes based on the mode of artificial counting at present, and namely number is got 1,000 grain (seed) and obtained grain weight by the weighing mode.Adopt the method for this artificial counting, because tired easily after the long intensive work of human eye, the counting accuracy is difficult to guarantee, and counting efficiency is lower.In order to improve counting efficiency, method and technology that various employing recognition technologies are counted grain have appearred at present, and for example following patent discloses method or the equipment that adopts recognition technology that grain (particle) is counted.
Application number is " 02126213.6 ", the patent of invention that name is called the method and apparatus of fast distinguishing particles " adopt scattered light histogram " discloses unique method and the equipment of the microscopic particles of distinguishing the protozoa that is suspended in liquid or the gas and other microorganisms and so on fast, this method comprises the particle that the intense light source irradiation of employing laser and so on will detect, by one group of light sensors scattered light around detection zone, convert detected light to electric signal, the signal that adopts at least one occurrence frequency/probability histogram relatively to derive, thus the microscopic particles that exists is carried out qualitative and/or quantitative identification.This patented method is mainly used in the microscopic particles that identification is suspended in protozoan in liquid or the gas and other microorganism and so on, identification has certain advantage for the suspended particulates under the three-dimensional environment, but this method as the deficiency of the identification existence of the non-suspended particle in plane is: (1) utilizes the light sensing equipment to have high requirements for equipment itself and experimental situation, and apparatus expensive is unfavorable for industrial applications; (2) utilize the light scattering technique generally need be by a plurality of angle shots reaching the location of particle for the particle recognition in the stereo-picture, and for the particulate identification of plane picture, this method also be not suitable for; (3) this method only relates to the identification of particulate (particulate group), and does not segment consideration to piling up with the adhesion situation.
Application number has just been announced name for the patent of invention of " 201110267255.0 " and has been called " automatic division method of graininess object in a kind of digital picture ", this method is at digital picture, especially characteristics such as the gray scale of graininess object, structure distribution and geometric configuration in the micro-image are used the automatic threshold method earlier with target and background separation; Calculate its gradient vector field then, searched key point in gradient vector field, desirable key point all has corresponding gradient vector to distribute at 8 neighborhoods, and its Grad is zero, and the key point of obtaining is as the center of each graininess object; Then define a new effective energy function based on gray scale and locus in order to the calculated direction gradient, it is replaced traditional shade of gray; Use active contour model at last and search the border of graininess object.This patented method still proposes high requirement for environment and the image taking environment of identifying object, namely this method is only applicable to the particle identification in the MIcrosope image, the general background of MIcrosope image is single, and has fluorescence labeling as the identification target, therefore with common shooting environmental under the particle identification of the non-MIcrosope image that collects notable difference is arranged, and whether be applicable to that there is certain query in this application scenarios.
Application number is called " cereal grain color sorting apparatus " for the patent of invention of " 95102156.7 " discloses name, the color foreign matter identical or transparent with grain also identified and remove to the utility model by both having detected and removed the color foreign matter different with grain with a sorter in visible-range near infrared range.This patent only relates to physical color identification method for sorting, does not propose solution for the counting of grain, and namely this method has certain advantage for the Quality Detection of grain, but identifies not practical use for the counting of grain.In addition, this method need be suitable near infrared gear and detect, and equipment cost and human cost that the cost of this equipment relates in making and detecting increase greatly.Therefore, this method is also inapplicable for the mensuration of grain counting and mass of 1000 kernel.
Application number is called " based on the granular material sorting classifying method of visual identity " for the patent of invention of " 200810052381.2 " discloses name, and it comprises and obtains sample image, image feature information pre-service, design category device, sets up characteristic information experts database and material sorting to be selected and characteristic information experts database and mate, carry out sorting classifying.This method is applicable to the sizing screening of granule materials, and therefore the enumeration problem for granule materials does not provide solution.Simultaneously, this method relates to relatively loaded down with trivial details flow process, defines the scope of application and the application scenarios of system to a certain extent.
In sum, though present existing technology has improved accuracy and counting efficiency to grain count, realized the robotization detection, but these technology need specialty and expensive equipment (such as laser or near infrared) usually, perhaps picture quality is had relatively high expectations, the structure more complicated is had relatively high expectations to operating personnel simultaneously.Be not suitable for light-weighted solution in the site work.
The utility model content
The utility model provides a kind of particle counting device at the deficiency that exists in the existing particle identification method of counting.
The utility model is achieved by following technical solution.
A kind of particle counting device, comprise transparent pallet and place transparent pallet material lighting source all around, above transparent pallet, be placed with camera installation, camera installation is connected with computing machine by data line, by the device drives of interface interchange camera installation, use keyboard and mouse to handle and take pictures.
The angle of described material lighting source can be regulated.
Described material lighting source is the adjustable White LED light source of light intensity.
Described material lighting source comprises one group of left side lamp that places transparent pallet left side, place the starboard light on transparent pallet right side and place the back of the body lamp of transparent tray back.
Adopt said apparatus to the method for particle counting to be, utilize camera arrangement to carry out the particle image acquisition, use intelligent image processing and machine learning techniques to realize the identification of particle is counted, when taking pictures, carry out image data management and processing by counting module, carry out fast image when finishing and identify in advance taking pictures, filter out undesirable photo, finish accurate counting, its main method step is as follows:
(1) particle image acquisition by the device drives of interface interchange camera, is used keyboard and mouse to handle and is taken pictures;
(2) the particle image of gathering is converted into the RGB numerical matrix, regulates the light and shade of image, strengthen contrast;
(3) utilize K-Means clustering algorithm automatic screening effective coverage;
(4) utilize the edge recognition technology to identify independently particle;
(5) individual particles that identifies is carried out modeling, obtain the contour shape feature about particle, the statistical model of size characteristic;
(6) estimate the adhesion piece of suspection, simulate filling according to the individual particles model, estimate the number that comprises, and output adhesion rule and cutting effect figure;
(7) the cutting effect figure to output manually passes judgment on, calculate the precision of identification in conjunction with the artificial counting result, the method of using machine learning is to identification and the counting of the particle under this situation of mode secondary computer systematic learning of the zone employing artificial counting that has adhesion and pile up, if the accuracy rate of identification counting is greater than 95%, finish the counting to particle, and output count results, otherwise, the result of artificial cognition is replaced the adhesion piece of estimating, again training pattern reaches stable up to the machine learning system model.
The particle image acquisition comprises the steps: in the above-mentioned steps (1)
A, aerosol sample taken out be positioned on transparent panel or the blank;
B, tiling particle make particle be covered with the viewfinder range of camera installation equably;
C, collection image;
D, repetition b and c step twice obtain n*3 and open image to reach the purpose at random of sampling, and n is sample size.
The beneficial effects of the utility model are:
Compared with prior art, adopt counting assembly described in the utility model, use intelligent image processing and machine learning techniques to realize the identification quick and precisely of particle is counted, by particle (grain) image that obtains with the particle image acquisition as handling object, at first utilize the image recognition counting to choose automatically for particle (grain) overlay area, and directly count for the grain in some non-complex zones (zone that does not have adhesion and pile up).For the zone that has adhesion and pile up, carry out Intelligent Recognition and counting by the method for machine learning.
The utility model only need adopt the camera arrangement of a simple and easy installation of energy and common family expenses camera that a collection of particle is taken pictures, and then image file is imported computing machine, and use is set up model image is handled.Adopt image recognition technology accurately to identify independently particle, utilize statistical model at the particle of dissimilar particle (Chinese sorghum, paddy rice etc.) identification and estimation adhesion, obtain precise counting at last.And use the mode of artificial counting that the result is detected, according to correction result training pattern again, constantly promote the precision of model.The camera arrangement that the utility model adopts is installed and is disposed easily, can be quick installed at multiple environment.Requirement to place light source and photographic equipment is not high yet, can obtain the photographic intelligence of particle fast.By statistical model and the machine learning method that has manual intervention, detection and training can be combined, increased the precision of computing machine identification.Can be applied to the production environment of particle counting easily, replace the mode of artificial counting, increase work efficiency significantly.The utility model has been given full play to fast, the advantage of strong, the batch processing of consistance as a result of the processing speed that modern image handles, thus realize to particle efficiently, counting quickly and accurately.Whole mensuration process is low to artificial dependence, shows good repeatability and high efficiency, and camera arrangement is not high to the quality requirements of image simultaneously, can extensively promote the use of in the particle counting field.
Description of drawings
Fig. 1 is the utility model structural representation;
Fig. 2 is for using the utility model to the process flow diagram of particle counting.
Among the figure: the transparent pallet of 1-, 2-left side lamp, 3-carries on the back lamp, 4-starboard light, 5-camera installation, 6-computing machine, 7-data line, 8-particle.
Embodiment
Further describe the technical solution of the utility model below in conjunction with accompanying drawing, but that claimed scope is not limited to is described.
As shown in Figure 1, a kind of particle counting device described in the utility model, comprise transparent pallet 1 and place transparent pallet 1 material lighting source all around, above transparent pallet 1, be placed with camera installation 5, camera installation 5 is connected with computing machine 6 by data line 7, by the device drives of interface interchange camera installation 5, use keyboard and mouse to handle and take pictures.
The angle of described material lighting source can be regulated.
Described material lighting source is the adjustable White LED light source of light intensity.
Described material lighting source comprises one group of left side lamp 2 that places transparent pallet 1 left side, place the starboard light 4 on transparent pallet 1 right side and place the back of the body lamp 3 at transparent pallet 1 back side.
As shown in Figure 2, a kind of particle counting method, utilize camera arrangement to carry out the particle image acquisition, use intelligent image processing and machine learning techniques to realize the identification of particle is counted, when taking pictures, carry out image data management and processing by counting module, carry out fast image when finishing and identify in advance taking pictures, filter out undesirable photo, finish accurate counting, its main method step is as follows:
(1) particle image acquisition by the device drives of interface interchange camera, is used keyboard and mouse to handle and is taken pictures;
(2) the particle image of gathering is converted into the RGB numerical matrix, regulates the light and shade of image, strengthen contrast;
(3) utilize K-Means clustering algorithm automatic screening effective coverage;
K-Means algorithm described in this step is input cluster number k, and the database that comprises n data object, and k cluster of variance minimum sandards satisfied in output.The k-means algorithm is accepted input quantity k; Then n data object is divided into k cluster in order to make the cluster that obtains satisfy, the object similarity in the same cluster is higher; And the object similarity in the different clusters is less.The cluster similarity is to utilize the average of object in each cluster to obtain a center object (center of attraction) to calculate.
(4) utilize the edge recognition technology to identify independently particle;
Edge recognition technology described in this step is to extract the peripheral profile of particle (grain) by the outline recognizer of image, and obtaining intermediate treatment object particle (grain) zone (namely not comprising background blank zone), its algorithm comprises Image outline identification based on wavelet transformation, based on the Image outline identification of the robert factor, based on the Image outline identification of the laplace operator factor, based on the profile identification of gaussian filtering.
(5) individual particles that identifies is carried out modeling, obtain the contour shape feature about particle, the statistical model of size characteristic;
(6) estimate the adhesion piece of suspection, simulate filling according to the individual particles model, estimate the number that comprises, and output adhesion rule and cutting effect figure;
(7) the cutting effect figure to output manually passes judgment on, calculate the precision of identification in conjunction with the artificial counting result, the method of using machine learning is to identification and the counting of the particle under this situation of mode secondary computer systematic learning of the zone employing artificial counting that has adhesion and pile up, if the accuracy rate of identification counting is greater than 95%, finish the counting to particle, and output count results, otherwise, the result of artificial cognition is replaced the adhesion piece of estimating, again training pattern reaches stable up to the machine learning system model.
The particle image acquisition comprises the steps: in the above-mentioned steps (1)
A, aerosol sample taken out be positioned on transparent panel or the blank;
B, tiling particle make particle be covered with the viewfinder range of camera installation equably;
C, collection image;
D, repetition b and c step twice obtain n*3 and open image to reach the purpose at random of sampling, and n is sample size.
Counting module described in the utility model can be realized the management processing to view data, the permission user gathers by the visualization interface management and obtains image, and can realize the identification of particle (grain) image is counted by image recognition and machine learning method.Machine Learning Theory mainly is design and analyzes some and allow the computing machine algorithm of " study " automatically.Machine learning algorithm is that a class is analyzed the acquisition rule automatically from data, and the algorithm that utilizes rule that unknown data is predicted.Because related to a large amount of statistical theories in the learning algorithm, machine learning and system of statistical inference student's federation are particularly close, are also referred to as Statistical Learning Theory.Therefore, the machine learning model of method described in the utility model by training can be identified intelligently and count for the particle that has adhesion and pile up (grain) zone.
When use at the utility model place, place a transparent pallet 1 at the photograph zone level, above transparent pallet 1, the left side, right side and the back side respectively arrange one group of angle and the adjustable material lighting source of light intensity, the material lighting source is respectively the left side lamp 2 that places transparent pallet 1 left side, place the starboard light 4 on transparent pallet 1 right side and place the back of the body lamp 3 at transparent pallet 1 back side, and is the high brightness White LED light source.Camera installation 5 is positioned over transparent pallet 1 top.When needs are taken pictures, particle 8 is interspersed among on the transparent pallet 1, and applied particle 8 at random.Camera installation 5 is connected to computing machine 6 by data line 7, by the device drives of interface interchange camera installation 5, uses keyboard and mouse to handle and take pictures.When taking pictures, carry out image data management and processing by counting module, carry out fast image when finishing and identify in advance taking pictures, filter out undesirable photo.According to each testing result training pattern and the parameter of adjustment model, constantly promote the precision of model.
If the utility model does not possess above device in use, can also replace in the following manner: particle (grain) sample is tiled on the flat board of white (or Transparent color), set up image capture device in about 50 centimeters of dull and stereotyped vertical direction, guarantee stable collection and the counting that can finish the particle image of light source and image capture device focal length, simple in structure, easy to operate.
Claims (4)
1. particle counting device, it is characterized in that: comprise transparent pallet (1) and place transparent pallet (1) material lighting source all around, be placed with camera installation (5) in transparent pallet (1) top, camera installation (5) is connected with computing machine (6) by data line (7), by the device drives of interface interchange camera installation (5), use keyboard and mouse to handle and take pictures.
2. particle counting device according to claim 1, it is characterized in that: the angle of described material lighting source can be regulated.
3. particle counting device according to claim 1, it is characterized in that: described material lighting source is the adjustable White LED light source of light intensity.
4. particle counting device according to claim 1 is characterized in that: described material lighting source comprises one group of left side lamp (2) that places transparent pallet (1) left side, place the starboard light (4) on transparent pallet (1) right side and place the back of the body lamp (3) at transparent pallet (1) back side.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201320172524 CN203177944U (en) | 2013-04-08 | 2013-04-08 | Particle counting apparatus |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201320172524 CN203177944U (en) | 2013-04-08 | 2013-04-08 | Particle counting apparatus |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN203177944U true CN203177944U (en) | 2013-09-04 |
Family
ID=49074627
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 201320172524 Expired - Lifetime CN203177944U (en) | 2013-04-08 | 2013-04-08 | Particle counting apparatus |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN203177944U (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103226088A (en) * | 2013-04-08 | 2013-07-31 | 贵州茅台酒股份有限公司 | Particulate counting method and device thereof |
| CN105928597A (en) * | 2016-04-14 | 2016-09-07 | 吉林大学 | Thousand-grain weight counter for corn particles and counting and weighing method of counter |
| CN106991683A (en) * | 2017-03-27 | 2017-07-28 | 西安电子科技大学 | Local active contour image segmentation method based on intermediate truth degree measurement |
| CN109991134A (en) * | 2019-03-29 | 2019-07-09 | 苏州精濑光电有限公司 | A kind of dust fall detection device |
| WO2023074818A1 (en) * | 2021-10-27 | 2023-05-04 | 株式会社安川電機 | Weighing system, support control system, weighing method, and weighing program |
-
2013
- 2013-04-08 CN CN 201320172524 patent/CN203177944U/en not_active Expired - Lifetime
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103226088A (en) * | 2013-04-08 | 2013-07-31 | 贵州茅台酒股份有限公司 | Particulate counting method and device thereof |
| CN105928597A (en) * | 2016-04-14 | 2016-09-07 | 吉林大学 | Thousand-grain weight counter for corn particles and counting and weighing method of counter |
| CN105928597B (en) * | 2016-04-14 | 2018-09-11 | 吉林大学 | A kind of mass of 1000 kernel calculating instrument and its counting weighing technique for corn particle |
| CN106991683A (en) * | 2017-03-27 | 2017-07-28 | 西安电子科技大学 | Local active contour image segmentation method based on intermediate truth degree measurement |
| CN106991683B (en) * | 2017-03-27 | 2019-10-08 | 西安电子科技大学 | Local active contour image segmentation method based on intermediate truth degree measurement |
| CN109991134A (en) * | 2019-03-29 | 2019-07-09 | 苏州精濑光电有限公司 | A kind of dust fall detection device |
| CN109991134B (en) * | 2019-03-29 | 2020-10-16 | 苏州精濑光电有限公司 | Dust fall detection equipment |
| WO2023074818A1 (en) * | 2021-10-27 | 2023-05-04 | 株式会社安川電機 | Weighing system, support control system, weighing method, and weighing program |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103226088B (en) | Particulate counting method | |
| CN203177944U (en) | Particle counting apparatus | |
| Sarrafzadeh et al. | Selection of the best features for leukocytes classification in blood smear microscopic images | |
| CN108171184A (en) | Method for distinguishing is known based on Siamese networks again for pedestrian | |
| CN102044069B (en) | Method for segmenting white blood cell image | |
| CN103984939B (en) | A kind of sample visible component sorting technique and system | |
| CN104008399B (en) | The recognition methodss of the gauge pointer shake based on support vector machine in a kind of instrument detection | |
| CN109684906B (en) | Method for detecting red fat bark beetles based on deep learning | |
| CN103984953A (en) | Cityscape image semantic segmentation method based on multi-feature fusion and Boosting decision forest | |
| CN104198324B (en) | Computer vision-based method for measuring proportion of cut leaves in cut tobacco | |
| CN104392240A (en) | Parasite egg identification method based on multi-feature fusion | |
| JP2016507759A5 (en) | ||
| CN101441721A (en) | Device and method for counting overlapped circular particulate matter | |
| CN103345617A (en) | Method and system for recognizing traditional Chinese medicine | |
| CN105095848B (en) | The method and system of target identification | |
| CN107103608B (en) | A kind of conspicuousness detection method based on region candidate samples selection | |
| US10769432B2 (en) | Automated parameterization image pattern recognition method | |
| CN103413124A (en) | Method for detecting round traffic sign | |
| CN103177266A (en) | Intelligent stock pest identification system | |
| Hobson et al. | Classifying anti-nuclear antibodies HEp-2 images: A benchmarking platform | |
| CN106407915A (en) | SVM (support vector machine)-based face recognition method and device | |
| CN107730499A (en) | A kind of leucocyte classification method based on nu SVMs | |
| CN108257124A (en) | A kind of white blood cell count(WBC) method and system based on image | |
| CN102938053A (en) | Sugarcane characteristic extraction and recognition method based on computer vision | |
| CN109115775A (en) | A kind of betel nut level detection method based on machine vision |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CX01 | Expiry of patent term |
Granted publication date: 20130904 |
|
| CX01 | Expiry of patent term |