CN104865198A - Machine vision-based fresh fat liver quality automatic grading system and method - Google Patents
Machine vision-based fresh fat liver quality automatic grading system and method Download PDFInfo
- Publication number
- CN104865198A CN104865198A CN201510204188.6A CN201510204188A CN104865198A CN 104865198 A CN104865198 A CN 104865198A CN 201510204188 A CN201510204188 A CN 201510204188A CN 104865198 A CN104865198 A CN 104865198A
- Authority
- CN
- China
- Prior art keywords
- foie gras
- fresh
- fresh foie
- grade
- quality automatic
- 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
Landscapes
- Sorting Of Articles (AREA)
Abstract
The invention discloses a machine vision-based fresh fat liver quality automatic grading system and method, the machine vision-based fresh fat liver quality automatic grading system includes a fresh fat liver acquisition platform, a dark box with a light source and an industrial camera, and a fresh fat liver quality automatic grading module, the light source in the dark box provides a uniform light illumination condition for fresh fat liver image information acquisition, the industrial camera in the dark box is located just above the fresh fat liver acquisition platform, and is used for fresh fat liver image information acquisition, the output end of the industrial camera is connected with the fresh fat liver quality automatic grading module, and the fresh fat liver quality automatic grading module can automatically grade the fresh liver fat quality grade according to acquired fresh fat liver images. The machine vision-based fresh fat liver quality automatic grading system can replace corresponding manual grading operation links in fresh fat liver processing fields, and can fast, objectively and accurately automatically grade the fresh liver fat color grade and weight grade.
Description
Technical field
The present invention relates to detection and the judgement of fresh foie gras quality grade, belong to processing of farm products and detection field, especially a kind of fresh foie gras quality automatic grading system based on machine vision and method thereof.
Background technology
At present, in traditional fresh foie gras processing industry, manual method is usually adopted to carry out classification to the quality grade of fresh foie gras.The fresh foie gras quality grade of China mainly comprises organoleptic indicator's grade, weight grade, physical and chemical index grade etc.The organoleptic indicator of fresh foie gras mainly comprises the color and luster of fresh foie gras, elasticity, smell, damage are levied, texture etc., and wherein, the color and luster of fresh foie gras is topmost evaluation index.When fresh foie gras organoleptic indicator classification, normally to be graded personnel by specialty, at fresh foie gras processing site, assess color grade by visual inspection, and then levy with reference to the elasticity of fresh foie gras, smell, damage, final assessment goes out organoleptic indicator's grade of fresh foie gras.Current fresh foie gras organoleptic indicator grade, in most of the cases, mainly refers to the color grade of fresh foie gras.Above-mentioned ranking method is a kind of subjectivity, qualitative but not objective, quantitative evaluation method, and ranking process relies on the factors such as the sense organ of grading personnel, experience and psychology, and it is fair and just to be not only difficult to ensure, and efficiency is low, error is large.And for fresh foie gras weight grading, be by artificial Weighing method equally, according to Different Weight, fresh foie gras is divided into different weight grade.This manually weighing divides the method for fresh foie gras weight grade, adds the Manual operation link of fresh foie gras processing site, not only affects fresh foie gras product quality, and affects fresh foie gras processing efficiency.Generally do not carry out the physical and chemical index classification work of fresh foie gras at fresh foie gras processing site, physical and chemical index classification mainly by measuring its crude protein, crude fat, moisture and grease leaching rate to representative liver sample, thus judges respective level.In a word, the artificial grading work carrying out correlated quality grading index in fresh foie gras production scene is the way of a kind of subjectivity, poor efficiency, the requirement of fresh foie gras suitability for industrialized production cannot be met, therefore, be necessary to further investigate fresh foie gras quality grading method and technology, to developing a kind of objective, fresh foie gras quality automatic grading system and method accurately and efficiently, realizing the online harmless of fresh foie gras quality grade and detecting fast.
Summary of the invention
The object of the invention is to have when manually grading for fresh foie gras processing site the problems such as subjectivity is strong, accuracy rate is low and grading efficiency is low, propose a kind of fresh foie gras quality automatic grading system based on machine vision and the method thereof that can adapt to the requirement of fresh foie gras processing site.
Main technical schemes of the present invention is:
A kind of fresh foie gras quality automatic grading system based on machine vision and method thereof, it comprises fresh foie gras acquisition platform, with camera bellows and the fresh foie gras quality automatic classification module of light source and industrial camera, light source in camera bellows provides unified illumination condition for the collection of fresh foie gras image information, industrial camera in camera bellows is positioned at the collection for fresh foie gras image information directly over fresh foie gras acquisition platform, the output terminal of industrial camera and fresh foie gras quality automatic classification model calling, fresh foie gras quality automatic classification module carries out fresh foie gras color grade automatically to the fresh foie gras image collected, the classification work of weight grade.
The background color of fresh foie gras acquisition platform of the present invention is light blue, and when opening camera bellows light source, the fresh foie gras background image that industrial camera collects should be light blue.
Utilize a method for the fresh foie gras quality automatic grading system based on machine vision as above, comprise the following steps:
(1) manual grading skill of fresh foie gras quality grade and record: be a quality grade evaluation group with 3 people, subjective appreciation is carried out to fresh foie gras sample evidence national standard " NY 67-1988 ", being 4 grades carry out record with numerical value 1,2,3,4 by color grade artificial division, is that 4 grades also carry out record with numerical value 1,2,3,4 by weight grade according to the artificial division of fresh foie gras weight range.
(2) collection of fresh foie gras image: the fresh foie gras image (sample image) having been completed manual quality's classification by the industrial camera collection in camera bellows.
(3) foundation of fresh foie gras image data base: repeat step (1), step (2), set up fresh foie gras image data base, have the record of its corresponding mass grade in every width sample image, comprise color grade and weight grade.
(4) foundation of fresh foie gras color and luster hierarchy model: first utilize image processing techniques to be partitioned into fresh foie gras target image to each the width sample image in fresh foie gras image data base, then fresh foie gras target image is carried out to the extraction of color and luster parameter, by the mean value of fresh foie gras region R, G, B tri-color components in every fresh foie gras target image (
,
,
) and standard deviation (
,
,
) totally 6 characteristic parameters are as the quantitative description of fresh foie gras color grade.Finally utilize multiple linear regression analysis method, set up with these 6 Color characteristics parameters be input, color grade value is that the fresh foie gras color and luster exported returns hierarchy model.
(5) foundation of fresh foie gras weight grading model: first utilize image processing techniques to be partitioned into fresh foie gras target image to each the width sample image in fresh foie gras image data base, then fresh foie gras target image is carried out to the extraction of weight parameter, using the fresh foie gras region area S of every fresh foie gras target image and perimeter L totally 2 characteristic parameters as the quantitative description of fresh foie gras weight grade.Finally utilize multiple linear regression analysis method, set up with area S and perimeter L be input, weight grade is the fresh foie gras weighted regression hierarchy model exported.
(6) fresh foie gras quality grade automatic classification: be placed in by the fresh foie gras not carrying out manual quality on the fresh foie gras acquisition platform of camera bellows, gather fresh foie gras image by the industrial camera directly over camera bellows, is transferred to fresh foie gras quality grading module by image information.First fresh foie gras quality grading module carries out Iamge Segmentation to fresh foie gras image, obtain fresh foie gras target image, then 6 characteristic parameters of fresh foie gras color grade, 2 characteristic parameters of weight grade are extracted, the color and luster hierarchy model finally set up according to step (4) and step (5) and weight grading model, calculate color grade and the weight grade of fresh foie gras respectively, thus realize the automatic classification of fresh foie gras quality grade.
Useful benefit of the present invention:
The integrated machine vision technique of the present invention, statistical theory and machine word software engineering technology develop the fresh foie gras quality automatic grading system based on machine vision, to collecting fresh foie gras image carries out related digital image process operation, the correlated characteristic characteristic parameter of the fresh foie gras quality of automatic extraction calculated mass grade (color grade, weight grade), thus the automatic classification realized fresh foie gras, compensate for the deficiency such as subjectivity, poor efficiency when fresh foie gras processing site manual quality grades, substantially increase the accuracy of on-the-spot classification, objectivity and work efficiency.
The color of the light source used in camera bellows in the present invention, light source type, illumination structure and light source are to the illumination illumination of fresh foie gras acquisition platform, all be through well-designed and repetition test demonstration, unified stable photoenvironment can be provided for the high-quality collection of fresh foie gras image.
The industrial camera type used in camera bellows in the present invention, setting height(from bottom) and position are all through repetition test demonstration, and the light source in camera bellows can be coordinated to collect high-quality fresh foie gras image.
Fresh foie gras acquisition platform formula in the present invention is in conjunction with the light blue nylon taffeta particular design of stainless steel material adhesion, the background of the fresh foie gras image collected is distinguished to some extent with fresh foie gras target to the full extent, reduces the complexity of successive image process and feature extraction operation.This fresh foie gras acquisition platform freely can dismantle mobile, convenient replacement, and meets state food professional hygiene standard completely from being designed into selection.
The fresh foie gras color and luster hierarchy model set up in the present invention and weight grading model, when when illumination condition, camera type selecting, camera setting height(from bottom), background layout are consistent, can not modify to existing model completely, just be applicable to the production requirement of the fresh foie gras production scene of different scales, realize the on-line automatic classification to fresh foie gras quality grade.
accompanying drawing illustrates:
Fig. 1 is structural representation of the present invention.
specific embodiments:
Below in conjunction with drawings and Examples, the present invention is further illustrated.
As shown in Figure 1, a kind of fresh foie gras quality automatic grading system based on machine vision, it is by fresh foie gras acquisition platform, with the camera bellows of light source and industrial camera, fresh foie gras quality automatic classification module three part composition, light source wherein in camera bellows provides unified illumination condition for the collection of fresh foie gras image information, industrial camera in camera bellows is placed in the collection for fresh foie gras image directly over fresh foie gras acquisition platform, the output terminal of industrial camera and fresh foie gras quality automatic classification model calling, fresh foie gras quality automatic classification module carries out fresh foie gras color grade automatically to the fresh foie gras image collected, the calculating of weight grade.
The fresh foie gras acquisition platform used in this fresh foie gras quality automatic grading system adopts high-quality stainless steel material processing and fabricating, meets the hygienic standard of state food industry, both artistic and practical.Whole geese fatty liver acquisition platform independent design, unconnected with other modules of whole system, independently detachable movement, background color is replaceable.The background color of this platform is designed to light blue, through lot of experiment validation, getting this color is the fresh foie gras target and background that background colour better can distinguish in fresh foie gras image, therefore the light blue nylon taffeta having one deck conveniently to replace using the mode adhesion meeting food security on platform is as platform background.
Utilize a method for the fresh foie gras quality automatic grading system based on machine vision as above, comprise the following steps:
(1) manual grading skill of fresh foie gras quality grade and record: be a quality grade evaluation group with 3 people, subjective appreciation is carried out to fresh foie gras sample evidence national standard " NY 67-1988 ", being 4 grades carry out record with numerical value 1,2,3,4 by color grade artificial division, is that 4 grades also carry out record with numerical value 1,2,3,4 by weight grade according to the artificial division of fresh foie gras weight range.
(2) collection of fresh foie gras image: the fresh foie gras image (sample image) having been completed manual quality's classification by the industrial camera collection in camera bellows.
(3) foundation of fresh foie gras image data base: repeat step (1), step (2), set up fresh foie gras image data base, have the record of its corresponding mass grade in every width sample image, comprise color grade and weight grade.
(4) foundation of fresh foie gras color and luster hierarchy model: first utilize image processing techniques to be partitioned into fresh foie gras target image to each the width sample image in fresh foie gras image data base, then fresh foie gras target image is carried out to the extraction of color and luster parameter, by the mean value of fresh foie gras region R, G, B tri-color components in every fresh foie gras target image (
,
,
) and standard deviation (
,
,
) totally 6 characteristic parameters are as the quantitative description of fresh foie gras color grade.Finally utilize multiple linear regression analysis method, set up with these 6 Color characteristics parameters be input, color grade value is that the fresh foie gras color and luster exported returns hierarchy model.
(5) foundation of fresh foie gras weight grading model: first utilize image processing techniques to be partitioned into fresh foie gras target image to each the width sample image in fresh foie gras image data base, then fresh foie gras target image is carried out to the extraction of weight parameter, using the fresh foie gras region area S of every fresh foie gras target image and perimeter L totally 2 characteristic parameters as the quantitative description of fresh foie gras weight grade.Finally utilize multiple linear regression analysis method, set up with area S and perimeter L be input, weight grade is the fresh foie gras weighted regression hierarchy model exported.
(6) fresh foie gras quality grade automatic classification: be placed in by the fresh foie gras not carrying out manual quality on the fresh foie gras acquisition platform of camera bellows, gather fresh foie gras image by the industrial camera directly over camera bellows, is transferred to fresh foie gras quality grading module by image information.First fresh foie gras quality grading module carries out Iamge Segmentation to fresh foie gras image, obtain fresh foie gras target image, then 6 characteristic parameters of fresh foie gras color grade, 2 characteristic parameters of weight grade are extracted, the color and luster hierarchy model finally set up according to step (4) and step (5) and weight grading model, calculate color grade and the weight grade of fresh foie gras respectively, thus realize the automatic classification of fresh foie gras quality grade.
The part that the present invention does not relate to prior art that maybe can adopt all same as the prior art is realized.
Claims (3)
1. the fresh foie gras quality automatic grading system based on machine vision and method thereof, it is characterized in that, comprise fresh foie gras acquisition platform, with the camera bellows of light source, industrial camera and fresh foie gras quality automatic classification module, quick, objective, exactly automatic classification is carried out to fresh foie gras color grade and weight grade.
2. according to claim 1, described a kind of fresh foie gras quality automatic grading system based on machine vision and method thereof, it is characterized in that, light source in camera bellows provides unified illumination condition for the collection of fresh foie gras image information, industrial camera in camera bellows is positioned at the collection for fresh foie gras image information directly over fresh foie gras acquisition platform, the output terminal of industrial camera and fresh foie gras quality automatic classification model calling, fresh foie gras quality automatic classification module carries out the classification work of fresh foie gras color grade, weight grade automatically to the fresh foie gras image collected.
3. according to claim 1, described a kind of fresh foie gras quality automatic grading system based on machine vision and method thereof, it is characterized in that, the background color of acquisition platform is light blue, when opening camera bellows light source, the fresh foie gras background image that industrial camera collects should be light blue.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510204188.6A CN104865198A (en) | 2015-04-28 | 2015-04-28 | Machine vision-based fresh fat liver quality automatic grading system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510204188.6A CN104865198A (en) | 2015-04-28 | 2015-04-28 | Machine vision-based fresh fat liver quality automatic grading system and method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN104865198A true CN104865198A (en) | 2015-08-26 |
Family
ID=53911181
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510204188.6A Pending CN104865198A (en) | 2015-04-28 | 2015-04-28 | Machine vision-based fresh fat liver quality automatic grading system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104865198A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106611428A (en) * | 2015-10-21 | 2017-05-03 | 中国人民解放军第三〇二医院 | Judgment card for quickly judging quality level of fructus schisandrae, and preparation method and application of judgment card |
| CN111480606A (en) * | 2020-04-23 | 2020-08-04 | 舟山国家远洋渔业基地科技发展有限公司 | Marine product grading treatment system for ocean fishing ship |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1603801A (en) * | 2004-11-02 | 2005-04-06 | 江苏大学 | Method and device for computer vision detection and classification of beef carcase quality |
| CN101561402A (en) * | 2009-05-07 | 2009-10-21 | 浙江大学 | Machine vision-based real-time detection and grading method and machine vision-based real-time detection and grading device for pork appearance quality |
| CN102156129A (en) * | 2009-12-02 | 2011-08-17 | 南京农业大学 | Beef quality intelligent grading system and method based on machine vision |
| CN103076288A (en) * | 2012-12-26 | 2013-05-01 | 中国科学院海洋研究所 | Automatic fish flesh grading device and method based on computer vision |
| CN103914708A (en) * | 2014-01-26 | 2014-07-09 | 冯平 | Food variety detection method and system based on machine vision |
-
2015
- 2015-04-28 CN CN201510204188.6A patent/CN104865198A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1603801A (en) * | 2004-11-02 | 2005-04-06 | 江苏大学 | Method and device for computer vision detection and classification of beef carcase quality |
| CN101561402A (en) * | 2009-05-07 | 2009-10-21 | 浙江大学 | Machine vision-based real-time detection and grading method and machine vision-based real-time detection and grading device for pork appearance quality |
| CN102156129A (en) * | 2009-12-02 | 2011-08-17 | 南京农业大学 | Beef quality intelligent grading system and method based on machine vision |
| CN103076288A (en) * | 2012-12-26 | 2013-05-01 | 中国科学院海洋研究所 | Automatic fish flesh grading device and method based on computer vision |
| CN103914708A (en) * | 2014-01-26 | 2014-07-09 | 冯平 | Food variety detection method and system based on machine vision |
Non-Patent Citations (8)
| Title |
|---|
| 中华人民共和国农业部: "《中华人民共和国国家标准》", 15 July 1988 * |
| 刘国敏 等: ""脐橙色泽与着色率的机器视觉检测技术研究"", 《江西农业大学学报》 * |
| 吴海娟 等: ""机器视觉技术在牛肉大理石花纹识别中的应用"", 《食品科学》 * |
| 孔彦龙 等: ""基于机器视觉的马铃薯质量和形状分选方法"", 《农业工程学报》 * |
| 杨东方 等: "《数学模型在生态学的应用及研究》", 1 May 2013, 北京：海洋出版社 * |
| 沈爱民 等: "《农产品质量安全与现代农业发展专家论坛论文集》", 31 December 2011, 北京：中国科学技术出版社 * |
| 逄滨 等: ""牛肉眼肌图像在线采集和实时分割系统"", 《农业机械学报》 * |
| 陈坤杰 等: ""牛肉自动分级技术研究进展分析"", 《农业机械学报》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106611428A (en) * | 2015-10-21 | 2017-05-03 | 中国人民解放军第三〇二医院 | Judgment card for quickly judging quality level of fructus schisandrae, and preparation method and application of judgment card |
| CN111480606A (en) * | 2020-04-23 | 2020-08-04 | 舟山国家远洋渔业基地科技发展有限公司 | Marine product grading treatment system for ocean fishing ship |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102156129B (en) | Beef quality intelligent grading system and method based on machine vision | |
| Zhou et al. | Strawberry maturity classification from UAV and near-ground imaging using deep learning | |
| CN104256882B (en) | Based on reconstituted tobacco ratio measuring method in the pipe tobacco of computer vision | |
| US8073235B2 (en) | Method and system for digital image analysis of ear traits | |
| CN104198324B (en) | Computer vision-based method for measuring proportion of cut leaves in cut tobacco | |
| CN104198325B (en) | Stem ratio measuring method in pipe tobacco based on computer vision | |
| CN102147402A (en) | Machine vision technology based method for rapidly detecting egg freshness | |
| CN109409441A (en) | Based on the coastal waters chlorophyll-a concentration remote sensing inversion method for improving random forest | |
| CN112669305B (en) | Metal surface rust resistance test bench and rust resistance evaluation method | |
| Rattanaphongphak et al. | Design of machine vision system for sugarcane buds or rings detection | |
| CN106780537A (en) | A kind of paper cocooning frame silk cocoon screening plant and method based on image procossing | |
| Marcon et al. | Estimation of total leaf area in perennial plants using image analysis | |
| CN202110131U (en) | Device for testing tobacco maturity | |
| CN112691939A (en) | Intelligent fruit and vegetable quality perception sorting robot system | |
| CN111881864A (en) | Wisdom farming crops growth full-period full-dynamic monitoring management system | |
| Smirnov et al. | Neural network for identifying apple fruits on the crown of a tree | |
| CN205158399U (en) | Multispectral vision formation of image da ye crops blade grading system | |
| CN104865198A (en) | Machine vision-based fresh fat liver quality automatic grading system and method | |
| CN103940751A (en) | Chilled meat quality evaluation method based on image processing technology | |
| CN116778343B (en) | Target image feature extraction method for comprehensive identification | |
| CN112858311A (en) | Rice phenotype monitoring system and method based on machine vision technology | |
| CN106326905A (en) | Multispectral vision imaging big-leaf crop leaf classification method and system | |
| CN111080103A (en) | Method for automatically evaluating crop yield | |
| CN109654559A (en) | A kind of healthy cooking system that can identify steam and oil smoke | |
| CN109622407B (en) | Image recognition-based walnut nondestructive testing equipment and method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| EXSB | Decision made by sipo to initiate substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20150826 |
|
| RJ01 | Rejection of invention patent application after publication |