CN1245630C - Agricultural and animal product nondestrctive detection method based on electronic visual sense and smell sense fusion technology and its device - Google Patents
Agricultural and animal product nondestrctive detection method based on electronic visual sense and smell sense fusion technology and its device Download PDFInfo
- Publication number
- CN1245630C CN1245630C CN 200410013836 CN200410013836A CN1245630C CN 1245630 C CN1245630 C CN 1245630C CN 200410013836 CN200410013836 CN 200410013836 CN 200410013836 A CN200410013836 A CN 200410013836A CN 1245630 C CN1245630 C CN 1245630C
- Authority
- CN
- China
- Prior art keywords
- smell
- agricultural
- data
- sense
- animal products
- 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 - Fee Related
Links
Images
Landscapes
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The present invention relates to a detection method aiming at agricultural and animal products. Images of agricultural and animal products, which are shot by a CCD camera, are transferred to a computer through an image collecting card; odors emitted from the agricultural and animal products under certain conditions (temperature, humidity and flow speed) are sucked to a reaction chamber and a gas sensor array through a miniature vacuum pump without oil so as to generate signals which are transferred to the computer through a conditioning circuit and an A/D collecting card. The computer simulates a human brain to carry out combine and pattern recognition processing to visual data and smell data; finally, the quality of the detected sample with different specifications, such as truth and falseness, quality, grades, qualified performance and unqualified performance, etc., is determined by the computer. The expansion to human eyes and noses can be realized through electronic vision and smell, and the outer characteristics, such as colors, shapes, textures, sizes, etc., of the agricultural and animal products, such as garden stuffs, meat, grains, beverages, etc. are combined with smell information, such as odors emitted in different periods; then, the expert knowledge in a knowledge base is combined with experience, and thus, the integrated discrimination is carried out.
Description
Technical field
The present invention relates to a kind of detection method, refer in particular to agricultural and animal products lossless detection method and device thereof based on electronics vision and sense of smell integration technology at agricultural and animal products.
Background technology
Owing to lack high-quality and high-caliber detection means, it is serious that China exports mixed level phenomenons such as agricultural and animal products mix, and is in totally unfavorable status in competition in the international market, and market share is constantly challenged.China is large agricultural country, and after the accession to WTO, the situation that agricultural and animal products face is more and more severe. artificial sense assessment method and conventional chemical analytical approach are still continued to use in the quality testing of present most agricultural and animal products.The conventional chemical analysis has higher accuracy and reliability, still, and the property consuming time of the pre-treatment of its sample, experiment itself and be again that many occasions institute is unallowed to the destructiveness of material.And subjective appreciation need be done some training very often usually, veteran expert finishes, judged result is along with the age, there is sizable individual difference in the difference of sex, recognition capability and spoken and written languages ability to express, even same personnel also produce different results with the variation of its condition and mood, be difficult to persist in reunification, objective standard, and labour intensity is big.Especially for short, the perishable agricultural and animal products of storage life, manual detection can not satisfy the requirement of whole detections far away.
Rising the eighties in 20th century abroad, the someone begins one's study based on the agricultural and animal products quality detection technology of electronics vision, main size, shape around agricultural and animal products, presentation quality features such as texture, color and surface imperfection are carried out. multiple fruit, vegetables such as apple, pears, capsicum, cucumber have been carried out Quality Detection and sorting, by retrieval, related U.S. patent is arranged, and the patent No. is: 5,732,147, patent name is; " Defective objectinspection and separation system using image analysis and curvaturetransformation (coming inspected object (apple) defective and the system distinguished) " with graphical analysis and curvature conversion.This invention detects the defective of body surfaces such as apple with image processing method and curvature conversion, and just powerless to the no abnormal and inner rotten apples in those surfaces.
Computing machine sense of smell technology is the technology that analysis, identification and the detection of complex of the novelty that grows up of the nineties smelt flavor and volatile ingredient.The foreign study persons use different gas sensor array, different analytical approachs respectively, and smell is differentiated and judged.Difference material such as carrot onion clearly, the smell that perfume, fillet etc. give out.By retrieval, relevant Chinese patent is arranged, application number: 03131660.3, based on the food smell fast non-destructive detection method and the device of gas sensor array technology; Other has related U.S. patent, the patent No. is: 6,450,008, patent name is: " Food applications of artificialolfactometry (be used for food manually smell the flavor detection method) ", more than invent the device that only food smell is detected, measured information is the information that food distributes the smell aspect also, and range of application is very limited.
Up to the present, what both at home and abroad the detection of most agricultural and animal products quality is adopted is the artificial sense assessment method, and also just rest on the research experiment stage based on the agricultural and animal products detection technique of single electronics vision or sense of smell. also more effective only according to the detection that visual information or sense of smell information are carried out some agricultural and animal products quality, but its detected quality information feature be after all have circumscribed, incomplete.For example the electronics vision can detect fruit color, shape, and external appearance characteristics such as size do not have significantly unusual rotten and the individuality of obvious peculiar smell arranged but can't detect outward appearance wherein.
Summary of the invention
In view of above-mentioned prior art development, purpose of the present invention is exactly to play to provide a kind of at lossless detection method and the device thereof of agricultural and animal products based on electronics vision and sense of smell integration technology. realize expansion by electronics vision and sense of smell to human eye and nose, with fruits and vegetables, the color of agricultural and animal products such as meat, cereal, beverage, shape, texture, appearance characteristics and its sense of smell information fusion such as smell of distributing such as size in various periods, and then with knowledge base in expertise and experience merge, carry out comprehensive distinguishing.
The objective of the invention is to realize by the following method;
At first set up knowledge base,,, set up database with the relevant conventional sense of this agricultural and animal products quality according to its examination criteria (national examination criteria, generic industry standard) to the agricultural and animal products of required mensuration.The image of these agricultural and animal products of taking with the CCD camera imports computing machine into through image pick-up card; And smell that these agricultural and animal products distribute and gas sensor array generation signal import computing machine into through modulate circuit, digital-to-analog circuit, the brain of computer mould personification is handled vision data and sense of smell data, merge and connect and carry out pattern recognition process with the database of setting up previously, thereby in computing machine, form the true and false, the quality that can determine sample, the knowledge base of grade, whether qualified different size quality.
Carry out test sample then:
1. during working sample, sample is inserted in the airtight daylighting chamber, pass in the computing machine by the picture of camera with sample;
2. image is handled, extracted eigenwert, allow sample in airtight daylighting chamber, distribute smell simultaneously;
3. in above-mentioned processing the smell of sample through solenoid valve, vacuum pump, flowmeter, with certain temperature, humidity, pressure, flow volume delivery to reaction chamber, with the gas sensor array sensor response; The flavor signal is smelt in generation, and delivers to computing machine, extracts individual features:
4. computing machine merges and mode treatment the characteristic signal that is extracted, and in conjunction with the knowledge base of having set up, provides the true and false, quality, grade, the whether qualified recognition result of sample.Recognition result shows by computing machine, the epicycle end of test (EOT):
5. after the preceding end of test (EOT), take out sample, feed pure air in daylighting chamber and reaction chamber, the smell that preceding once test is carried over blows away, and gas sensor is restored.
Described computing machine merges and mode treatment the characteristic signal that is extracted, wherein data fusion adopts conventional analysis of statistical data method (pattra leaves department technology, regretional analysis, fractal, principal component analysis (PCA), Fourier transform, isolated component method, fuzzy theory) and neural network and genetic algorithm when making up high-precision real pattern classification system handle vision and sense of smell data and connect and learn, train with database, obtain a knowledge base. make the system that is developed can determine the true and false of sample, whether qualified good and bad, grade be.
Described fusion is divided into raw data merges, the fusion of many levels such as characteristic merges, decision-making data fusion.Raw data merges and characteristic merges processing and the fusion that mainly comprises vision data and sense of smell data itself, in handling, vision data comprises general image processing method (background segment, filtering and noise reduction, rim detection, texture detects) and image characteristics extraction, and the sense of smell data processing comprises the sensing data smoothing denoising, feature extraction, normalized: the method for using in the raw data of vision data merges has independent component analysis, principal component analysis (PCA), Fourier transform, thresholding is cut apart, it mainly is Convolution Analysis that the sense of smell raw data merges, quadrature analysis, the isolated component method is analyzed; The characteristic fusion is meant extracts eigenwert earlier from each sensor, adopt discriminant analysis, neural network, regretional analysis, pattra leaves department technology, genetic algorithm, K nearest neighbor method to merge then on the basis of the eigenwert of being extracted; The fusion of decision data level be according to the judged result that electronics vision and sense of smell are provided use pattra leaves department method, fuzzy theory is judged, from many aspects such as the profile of survey agricultural and animal products, color, size, texture, smells to a comprehensive evaluation.
Described agricultural and animal products the cannot-harm-detection device based on electronics vision and sense of smell integration technology is made up of three parts: electronics vision system, electronics olfactory system, pattern-recognition and fused data disposal system, and electronics vision system and electronics olfactory system couple together by the daylighting chamber; The electronics vision system is by the CCD camera, image pick-up card, the daylighting chamber, light source, compositions such as light adjusting circuit, CCD camera wherein, light source, light adjusting circuit is fixed on the daylighting chamber interior, image pick-up card is fixed on computer-internal. and electronics sense of smell equipment comprises the gas sensor array that is linked in sequence by signal wire, the gas sensor modulate circuit, be contained in the A/D capture card of computer-internal, and reaction chamber, the power supply supplying module, control test environment and gas flow device, the power supply supplying module that is arranged in the next door is given gas sensor array respectively, gas sensor modulate circuit etc. provides the high precision D.C. regulated power supply. gas sensor array, reaction chamber is arranged in the thermostat, wherein control test environment and the gas flow device comprises thermostat, filtrator, flowmeter, gas piping, valve, there is not oily minipump etc. be disposed in order filtrator on the gas piping, there is not oily minipump, flowmeter, solenoid valve.Described thermostat is a constant temperature water bath apparatus, the temperature that is used for controling environment.
Described gas sensor array is made up of 8 to 16 dissimilar gas sensors, evenly alternative arrangement is in the reaction chamber of bar shaped, this bar shaped reaction chamber is arranged in the thermostatic box, inside surface is smooth not to have the gas dead angle, two ends are respectively equipped with air intake opening and gas outlet, be respectively equipped with admission line and outlet pipe on this air intake opening and the gas outlet: flowmeter has been linked in sequence on the admission line, there is not oily micro-vacuum pump, two-position three way magnetic valve, the inlet end of solenoid valve is connected with the two-way pipeline, wherein a pipeline is connected to the daylighting chamber, draw the smell that sample to be tested wherein distributes, make and carrying out Flame Image Process simultaneously, smell flavor and detect, another pipeline is connected to pure air sensor is restored usefulness.
Pressure, temperature, humidity sensor are housed in the described gas circuit, can make institute's test sample like this this produces head space gas under same pressure, temperature, humidity, the precision and the repeatability of test are improved: pressure, temperature, humidity sensor output is connected, and power supply is provided and solenoid valve is controlled.
During work, the CCD camera is taken the image of agricultural and animal products, import computing machine into through image pick-up card, and agricultural and animal products (temperature under certain condition, humidity, flow velocity) smell that distributes produces signal through having oily minipump sucting reaction chamber and gas sensor array, importing computing machine into through modulate circuit, A/D capture card. the brain of computer mould personification merges and pattern recognition process vision data and sense of smell data, determine the true and false of sample at last by computing machine, quality, the quality of grade, different size such as whether qualified.
The invention has the beneficial effects as follows:
Simulate the vision and the olfactory system of humans and animals based on the agricultural and animal products Non-Destructive Testing of electronics vision and sense of smell integration technology, what obtain is not the simple superposition of visual information and scent signal in the sample, but apish information fusion ability, vision and sense of smell information fusion are got up, pattern classification system is handled vision and sense of smell data during with high-precision real, and compared with information in the database of setting up through study, differentiate, sample quality is carried out comprehensive detection, thereby have artificial intelligence.Can be used for differentiating authenticity of products, control from the former whole process of production of expecting technology, thereby product quality is guaranteed.Based on electronics vision and sense of smell information fusion technology can be agricultural and animal products, food service industry provides new product lossless detection method and device, and is auxiliary or replace professional judge personnel with it.Also can be widely used in the monitoring of environment, tobacco business detects the total quality of tobacco, and medical department is made diagnosis or the like by comprehensive patient's complexion information, exhalation smell and the body odour information of distributing to the state of an illness.
Pattern classification system is handled vision and sense of smell data during with high-precision real, has improved the sensitivity of test, and selectivity and repeatability enlarge its identification range.Agricultural and animal products the cannot-harm-detection device of the present invention can not only be measured the trace contained in object color, shape, texture, size and the smell of surveying, trace and even ultratrace chemical constitution fast, especially can rapidly and accurately measurement data be converted to and the corresponding to result of expert's subjective appreciation. it not only can measure different signals according to various smell, and these signals and the signal in the knowledge base that study is set up can be compared, discern judgement.
The present invention compares with single vision or sense of smell detection technique, the information that obtains more comprehensively, its reliability, repeatability and adaptability are improved.Compare with the conventional chemical analytical approach, the method technical operation is fast and convenient, and sample does not need pre-treatment, does not also need any organic solvent to extract, and measures a sample less than 5 minutes, and the recognition reaction that unknown sample is had artificial intelligence.Compare with people's sense organ, measurement result is more objective, reliable.
The present invention introduces the hi-tech-integration technology in the information science field, electronics vision and sense of smell information fusion got up the agricultural and animal products quality is carried out comparatively comprehensively Non-Destructive Testing, to be used for agricultural and animal products Quality Detection and automatic classification process based on the high-new detection technique of computer technology, both can liberate the labour, the interference caused by subjective factors of getting rid of the people can be carried out the comprehensive evaluation of agricultural and animal products quality again quickly and accurately.Can carry out quick, easy, objective detection to processing, storage and the transportation of agricultural and animal products, accurately, in real time, the agricultural and animal products quality is guaranteed thereby effectively the agricultural and animal products production run is monitored.
Description of drawings
Fig. 1: technical scheme synoptic diagram of the present invention
Fig. 2; Electronics vision and sense of smell data fusion block diagram
Fig. 3: application example of the present invention (at apple) technology path synoptic diagram
Fig. 4: application example of the present invention is realized the hardware synoptic diagram
Fig. 5: control test environment and gas flow apparatus structure sketch in the application example of the present invention
Fig. 6: the data processing software interface of the embodiment of the invention
Fig. 7: the result that the embodiment of the invention detects red fuji apple
Among the figure: 1.CCD camera, 2. light adjusting circuit, 3. image pick-up card, 4. display, 5. computing machine, 6.A/D capture card, 7. gas sensor modulate circuit, 8. power supply supplying module, 9. gas sensor array, 10. reaction chamber, 11. control test environment and gas flow device, 12. the agricultural and animal products of surveying, daylighting chambers 13., 14. light sources, 15. filtrators, 16. thermostats, 17. solenoid valves, 18. vacuum pumps, 19. flowmeters, 20. Temperature Humidity Sensors, 21. pressure transducers
Embodiment
The present invention has versatility to the Non-Destructive Testing of agricultural and animal products, but because the agricultural and animal products kind is a lot, therefore the present invention is only for an embodiment who is used for red fuji apple, the detection of other agricultural and animal products can be with reference to the method for this embodiment, specifically at the evaluation criterion of the sample of being surveyed, set up a new knowledge base, just can test such agricultural and animal products.
Present embodiment is consulted Fig. 3, the system schema synoptic diagram that the present invention detects apple. select the apple of various quality grades earlier according to national standard, detection means is carried out grade estimation and classification routinely earlier, then with these apples as master sample, use based on the cannot-harm-detection device of electronics vision and sense of smell integration technology it is carried out Non-Destructive Testing, set up knowledge base.
Conventional sense means among the figure are fully by China's OB10651-89 fresh apple in 1993 grade scale.
Electronics vision among the figure comprises image acquisition, the image pre-service, graphical analysis and feature extraction, analyze (as the fruit conformal analysis among the figure, fruit face look and defect analysis, fruit body dimension analysis etc.), wherein image acquisition is by the CCD camera apple sample in the daylighting chamber to be taken, import computing machine into through image pick-up card, wherein daylighting chamber interior is furnished with light source, light adjusting circuit and rotary test platform etc., image acquisition circuit fastens in computer-internal, by computing machine to Stepping Motor Control, stepper motor driven rotary test board is rotated, and then drive the apple rotation, make camera can photograph the apple whole surface image.
The image pre-service comprises filtering and noise reduction, softwares such as apple image and background segment.The fruit conformal analysis extracts the marginal information of apple and carries out fourier expansion by image processing method, with the first few items coefficient of fourier expansion formula the shape of apple is judged, draws its circularity, flexibility, shape facilities such as symmetry.Fruit body dimension analysis with the maximum transverse diameter of computer software extraction apple, draws the actual maximum transverse diameter of apple then by homing method.
Electronics sense of smell among the figure comprises gas sensor array and apple odor response device and gas sensor array data acquisition, data pre-service, feature extraction, analyze, being placed on the indoor apple of daylighting distributes smell and is drawn by oilless vacuum pump, through piping and solenoid valve, shake the reaction of device array with certain flow by sensor array reaction chamber and gas biography and produce signal.This signal imports computing machine into through sensor modulate circuit and A/D capture card.
The pre-service of gas sensor array data comprises processing such as smothing filtering, base standardization, normalization, feature extraction is that some can represent the data of this course of reaction information from resulting extracting data, as temporal signatures (reaction maximal value, stationary value, mean value integrated value, differential value etc.), frequency domain character (fourier coefficient etc.), this feature is judged the olfactory characteristic of this apple as the input unit of neural network with neural network.
Information fusion among the figure, it at first is the raw data fusion in electronics sense of smell or electronics vision, base standardization after the sensor array data acquisition, normalized and electronics vision obtain the not raw data fusion of the image of ipsilateral of apple, as using the not image information removal background of ipsilateral of thresholding split plot design comprehensive three secondary apples, with the independent component analysis method the noise remove of image and gas sensor, and visual signal and scent signal carried out Fourier transform, obtain feature with principal component analysis (PCA).Carrying out characteristic then on image characteristics extraction and gas sensor array feature extraction basis merges, obtain apple fruit footpath size with regretional analysis, fractal characteristic to the color of apple carries out discriminant analysis, neural network obtains the apple color grade, obtains the shape class of apple with pattra leaves department's technology and analysis of neural network: with principal component analysis (PCA), genetic algorithm to gas sensor
Characteristic merges the olfactory characteristic that obtains apple. and on the basis that obtains apple visual signature and olfactory characteristic, carry out decision level fusion at last, obtain the final mass grade of apple with pattra leaves department method, fuzzy theory.
The hardware unit synoptic diagram of the embodiment of the invention as shown in Figure 4, electronics vision system and electronics olfactory system couple together by daylighting chamber (13); The electronics vision is made up of CCD camera (1), image pick-up card (3), daylighting chamber (13), light source (14), light adjusting circuit (2) etc., wherein CCD camera (1), light source (14), light adjusting circuit (2) are fixed on inside, daylighting chamber (13), and image pick-up card (3) is fixed on computer-internal (5).
Electronics sense of smell equipment comprises gas sensor array (9), gas sensor modulate circuit (7), be contained in the A/D capture card (6) of computer-internal, reaction chamber (10), power supply supplying module (8) control test environment and gas flow device (11) etc., the structure of wherein controlling test environment and gas flow device (11) as shown in Figure 5, daylighting chamber (13) is placed in the thermostat (16), it is a constant temperature water bath apparatus, the temperature that is used for controling environment.Be furnished with Temperature Humidity Sensor (20), pressure transducer (21) in thermostat (16), can make institute's test sample like this this produces head space gas under same pressure, temperature, humidity, and the precision of test and repeatability are improved.Be disposed in order filtrator (15) on the gas piping, solenoid valve (17), do not having oily minipump (18), flowmeter (19).
Described gas sensor array (9) is made up of 12 dissimilar gas sensors, evenly alternative arrangement is in the reaction chamber of bar shaped, this bar shaped reaction chamber is arranged in the thermostatic box, inside surface is smooth not to have the gas dead angle, two ends are respectively equipped with air intake opening and gas outlet, be respectively equipped with admission line and outlet pipe on this air intake opening and the gas outlet: the flowmeter that has been linked in sequence on the admission line (10), there is not oily micro-vacuum pump (18), two-position three way magnetic valve (17), the inlet end of solenoid valve is connected with the two-way pipeline, wherein a pipeline is connected to daylighting chamber (13), draw the smell that sample to be tested wherein distributes, make and carrying out Flame Image Process simultaneously, smell flavor and detect, another pipeline is connected to pure air sensor is restored usefulness.
Each gas sensor (9) and daylighting chamber (13) indoor pressure transducer (21), Temperature Humidity Sensor (20) exported and is connected in described modulate circuit plate (7) and the bar shaped gas sensor array reaction chamber (10), and power supply is provided and solenoid valve (17) is controlled.During work, CCD camera (1) is taken the image (12) of apple, import computing machine (5) into through image pick-up card (3), and apple (temperature under certain condition, humidity, flow velocity) smell that distributes does not produce signal through there being oily minipump (18) sucting reaction chamber (10) with gas sensor array (9), imports computing machine into through modulate circuit (7), A/D capture card (6).
The brain of computer mould personification merges and pattern recognition process vision data and sense of smell data, at last by the true and false, the quality of computing machine decision sample, and the quality of grade, different size such as whether qualified.
Above electronics sense of smell and electronics vision are carried out simultaneously, and its software interface is as shown in Figure 6. and finally obtain the quality grade of apple, as shown in Figure 7, comprise smell, the outward appearance of apple, size, a comprehensive evaluation of color etc.
Claims (7)
1. based on the agricultural and animal products lossless detection method of electronics vision and sense of smell integration technology, it is characterized in that:
At first set up knowledge base, agricultural and animal products to required mensuration, according to its examination criteria, set up database with the relevant conventional sense of this agricultural and animal products quality, the image of these agricultural and animal products of taking with the CCD camera imports computing machine into through image pick-up card, and smell that these agricultural and animal products distribute and gas sensor array produce signal through modulate circuit, digital-to-analog circuit imports computing machine into, the brain of computer mould personification is handled vision data and sense of smell data, merge and connect and carry out pattern recognition process, thereby in computing machine, form the true and false that can determine sample with the database of setting up previously, good and bad, grade, whether qualified, the knowledge base of different size quality;
Carry out the test sample of following steps then:
1. during working sample, sample is inserted in the airtight daylighting chamber, pass in the computing machine by the picture of camera with sample;
2. image is handled, extracted eigenwert, allow sample in airtight daylighting chamber, distribute smell simultaneously;
3. in above-mentioned processing the smell of sample through solenoid valve, vacuum pump, flowmeter, with certain temperature, humidity, pressure, flow volume delivery to reaction chamber, with the gas sensor array sensor response; The flavor signal is smelt in generation, and delivers to computing machine, extracts individual features:
4. computing machine merges and mode treatment the characteristic signal that is extracted, and provides the true and false, quality, grade of sample, whether qualified recognition result, and recognition result shows by computing machine, the epicycle end of test (EOT);
5. after the preceding end of test (EOT), take out sample, feed pure air in daylighting chamber and reaction chamber, the smell that preceding once test is carried over blows away, and gas sensor is restored.
2. the agricultural and animal products lossless detection method based on electronics vision and sense of smell integration technology according to claim 1, it is characterized in that described computing machine merges and mode treatment the characteristic signal that is extracted, pattern classification system was handled vision and sense of smell data and is connected with database and learns, trains when wherein data fusion adopted conventional analysis of statistical data method, neural network and genetic algorithm to make up high-precision real, made described system can determine the true and false, quality, the grade, whether qualified of sample.
3. the agricultural and animal products lossless detection method based on electronics vision and sense of smell integration technology according to claim 1 is characterized in that the fusion described in step 4. is divided into the fusion of raw data fusion, characteristic fusion, decision-making data fusion many levels; Raw data merges and characteristic merges processing and the fusion that comprises vision data and sense of smell data itself, in handling, vision data comprises general image processing method: background segment, filtering and noise reduction, rim detection, texture detects and image characteristics extraction, and the sense of smell data processing comprises the sensing data smoothing denoising, feature extraction, normalized: the method for using in the raw data of vision data merges has independent component analysis, principal component analysis (PCA), Fourier transform, thresholding is cut apart, and the fusion of sense of smell raw data comprises Convolution Analysis, quadrature analysis, the isolated component method is analyzed; The characteristic fusion is meant extracts eigenwert earlier from each sensor, adopt discriminant analysis, neural network, regretional analysis, pattra leaves department technology, genetic algorithm, K nearest neighbor method to merge then on the basis of the eigenwert of being extracted; The fusion of decision data level be according to the judged result that electronics vision and sense of smell are provided use pattra leaves department method, fuzzy theory is judged, from the profile of survey agricultural and animal products, color, size, texture, smell many aspects to a comprehensive evaluation.
4. realize the device of the described agricultural and animal products lossless detection method based on electronics vision and sense of smell integration technology of claim 1, it is characterized in that forming by three parts; Electronics vision system, electronics olfactory system, pattern-recognition and fused data disposal system, electronics vision system and electronics olfactory system couple together by daylighting chamber (13); The electronics vision system is made up of CCD camera (1), image pick-up card (3), daylighting chamber (13), light source (14), light adjusting circuit (2), wherein CCD camera (1), light source (14), light adjusting circuit (2) are fixed on inside, daylighting chamber (13), and image pick-up card (3) is fixed on computer-internal; The electronics olfactory system comprises the gas sensor array (9) that is linked in sequence by signal wire, gas sensor modulate circuit (7), be contained in the A/D capture card (6) of computer-internal, and reaction chamber (10), power supply supplying module (8), control test environment and gas flow device (11), the power supply supplying module (8) that is arranged in reaction chamber (10) next door provides the high precision D.C. regulated power supply for respectively gas sensor array (9), gas sensor modulate circuit (7), and gas sensor array (9), reaction chamber (10) are arranged in thermostat (16) lining; Wherein control test environment and gas flow device and comprise thermostat (16), filtrator (15), flowmeter (19), gas piping, solenoid valve (17), do not have oily minipump (18), be disposed in order filtrator (15) on the gas piping, do not have oily minipump (18), flowmeter (19), solenoid valve (17); Described gas sensor array (9) is made up of 8 to 16 dissimilar gas sensors, evenly alternative arrangement is in reaction chamber (10) lining of bar shaped, this bar shaped reaction chamber (10) is arranged in thermostat (16) lining, inside surface is smooth not to have the gas dead angle, two ends are respectively equipped with air intake opening and gas outlet, be respectively equipped with admission line and outlet pipe on this air intake opening and the gas outlet: the flowmeter that has been linked in sequence on the admission line (19), there is not oily micro-vacuum pump (18), solenoid valve (17), the inlet end of solenoid valve (17) is connected with the two-way pipeline, wherein a pipeline is connected to daylighting chamber (13), draw the smell that sample to be tested wherein distributes, make and carrying out Flame Image Process simultaneously, smell flavor and detect, another pipeline is connected to pure air and restores to be used for sensor; The data that electronics vision system, electronics olfactory system collect are transferred to pattern-recognition and fused data disposal system through lead.
5. device according to claim 4 is characterized in that thermostat (16) is a constant temperature water bath apparatus, the temperature that is used for controling environment.
6. device according to claim 4 is characterized in that being equipped with in the described daylighting chamber (13) pressure transducer (21), Temperature Humidity Sensor (20).
7. device according to claim 4, it is characterized in that each gas sensor is exported with pressure, temperature, humidity sensor in the gas sample generating chamber in described modulate circuit (7) and the bar shaped gas sensor array reaction chamber (10) is connected, and power supply is provided and solenoid valve is controlled.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 200410013836 CN1245630C (en) | 2004-01-08 | 2004-01-08 | Agricultural and animal product nondestrctive detection method based on electronic visual sense and smell sense fusion technology and its device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 200410013836 CN1245630C (en) | 2004-01-08 | 2004-01-08 | Agricultural and animal product nondestrctive detection method based on electronic visual sense and smell sense fusion technology and its device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1556412A CN1556412A (en) | 2004-12-22 |
| CN1245630C true CN1245630C (en) | 2006-03-15 |
Family
ID=34351113
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 200410013836 Expired - Fee Related CN1245630C (en) | 2004-01-08 | 2004-01-08 | Agricultural and animal product nondestrctive detection method based on electronic visual sense and smell sense fusion technology and its device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN1245630C (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101013079B (en) * | 2007-02-07 | 2010-11-10 | 浙江大学 | Small-sized material digitalized detecting and grading apparatus |
| CN101144780B (en) * | 2006-09-14 | 2011-05-11 | 北京工商大学 | Pork freshness intelligent detection device |
| CN101692053B (en) * | 2009-10-09 | 2012-01-18 | 江苏大学 | Multi-sensing information fusion based instrumental intelligent evaluation method for quality of famous tea |
Families Citing this family (26)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101074947B (en) * | 2007-06-27 | 2010-09-15 | 江苏大学 | Method and apparatus for inspecting two-kind gas sensor array combination odour |
| CN101234381B (en) * | 2008-03-07 | 2011-09-07 | 天津市华核科技有限公司 | Granular material sorting classifying method based on visual sense recognition |
| US20110172931A1 (en) | 2008-06-23 | 2011-07-14 | Atonarp Inc. | System for Handling Information Relating to Chemical Substances |
| CN101419213B (en) * | 2008-12-08 | 2012-11-07 | 浙江大学 | Detection method for bird egg crack by utilizing volatile matter |
| CN101936912A (en) * | 2010-08-25 | 2011-01-05 | 江苏大学 | Method and device for detecting freshness of fish based on olfaction visualization |
| DE102011007430B4 (en) * | 2011-04-14 | 2013-02-07 | Hauni Maschinenbau Ag | Promotion of rod-shaped articles of the tobacco processing industry |
| CN102209218A (en) * | 2011-05-20 | 2011-10-05 | 靳波 | Television set with fragrance |
| CN102636433B (en) * | 2012-02-17 | 2017-03-01 | 海尔集团公司 | A kind of food freshness detection means and the refrigerator using this device |
| CN102721793B (en) * | 2012-06-11 | 2015-04-22 | 江苏大学 | Method and device for digitally detecting quality of edible vinegar |
| CN102967597A (en) * | 2012-09-26 | 2013-03-13 | 江苏大学 | Olfactory imaging sensing technology based yellow wine storage time identification method and identification system |
| CN103976468A (en) * | 2014-04-30 | 2014-08-13 | 河南科技大学 | Tobacco leaf grading method |
| CN103940714B (en) * | 2014-05-13 | 2017-02-15 | 武汉大学 | Imitated artificial haze monitoring system and method |
| CN104049068B (en) * | 2014-06-06 | 2015-11-18 | 中国肉类食品综合研究中心 | The non-destructive determination device of fresh poultry meat freshness and assay method |
| CN105181912B (en) * | 2015-06-30 | 2017-09-05 | 江苏大学 | A kind of Noninvasive Measuring Method of Freshness in rice storage |
| CN106338517A (en) * | 2016-09-23 | 2017-01-18 | 江苏大学 | Intelligent judgment method for fruit freshness based on coordination of visual information and olfactory information |
| CN107886095A (en) * | 2016-09-29 | 2018-04-06 | 河南农业大学 | A kind of classifying identification method merged based on machine vision and olfactory characteristic |
| CN107121351B (en) * | 2017-03-28 | 2020-02-18 | 华南理工大学 | Acidic solution analysis method |
| CN107991460A (en) * | 2017-11-16 | 2018-05-04 | 吉林大学 | A kind of soil nutrient device for fast detecting based on bionic olfactory |
| WO2020089929A2 (en) | 2018-10-29 | 2020-05-07 | Tata Consultancy Services Limited | Apparatus and method for multimodal sensing and monitoring of perishable commodities |
| CN111220496A (en) * | 2020-03-11 | 2020-06-02 | 中国农业科学院农业信息研究所 | Apple quality detector and detection method |
| CN112036470A (en) * | 2020-08-28 | 2020-12-04 | 扬州大学 | Cloud transmission-based multi-sensor fusion cucumber bemisia tabaci identification method |
| CN116380814A (en) * | 2021-06-23 | 2023-07-04 | 长江师范学院 | Crisp plum maturity detection equipment based on visual identification |
| CN113390833A (en) * | 2021-06-23 | 2021-09-14 | 长江师范学院 | Crisp plum maturity judging method based on multiple detection modes |
| CN113607782B (en) * | 2021-07-28 | 2024-06-21 | 浙江工业大学 | Visual perception type water quality early warning system and method enhanced by olfactory signals |
| CN114047751A (en) * | 2021-10-28 | 2022-02-15 | 成都信息工程大学 | Robot three-dimensional inspection system and method in refrigerator |
| CN116577473B (en) * | 2023-07-14 | 2023-11-17 | 北京市农林科学院 | Detection method and device for strawberry mechanical damage occurrence time |
-
2004
- 2004-01-08 CN CN 200410013836 patent/CN1245630C/en not_active Expired - Fee Related
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101144780B (en) * | 2006-09-14 | 2011-05-11 | 北京工商大学 | Pork freshness intelligent detection device |
| CN101013079B (en) * | 2007-02-07 | 2010-11-10 | 浙江大学 | Small-sized material digitalized detecting and grading apparatus |
| CN101692053B (en) * | 2009-10-09 | 2012-01-18 | 江苏大学 | Multi-sensing information fusion based instrumental intelligent evaluation method for quality of famous tea |
Also Published As
| Publication number | Publication date |
|---|---|
| CN1556412A (en) | 2004-12-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN1245630C (en) | Agricultural and animal product nondestrctive detection method based on electronic visual sense and smell sense fusion technology and its device | |
| Huang et al. | Non-destructively sensing pork’s freshness indicator using near infrared multispectral imaging technique | |
| CN101251526B (en) | Method and apparatus for nondestructively testing food synthetic quality | |
| Khulal et al. | Intelligent evaluation of total volatile basic nitrogen (TVB-N) content in chicken meat by an improved multiple level data fusion model | |
| CN100495032C (en) | Method and apparatus for detecting surface quality of egg | |
| Kamruzzaman et al. | Introduction to hyperspectral imaging technology | |
| Pan et al. | Eggshell crack detection based on computer vision and acoustic response by means of back-propagation artificial neural network | |
| CN110443262B (en) | Computer vision-based rapid nondestructive testing method and device for freshness of fish body | |
| CN104198325B (en) | Stem ratio measuring method in pipe tobacco based on computer vision | |
| Wang et al. | A multimodal machine vision system for quality inspection of onions | |
| CN108663367A (en) | A kind of egg quality lossless detection method based on egg unit weight | |
| Chao et al. | On-line inspection of poultry carcasses by a dual-camera system | |
| CN102788794A (en) | Device and method for detecting pesticide residues on leaves of leaf vegetables on basis of multi-sensed information fusion | |
| CN104568824B (en) | Shrimps grade of freshness detection method and device based on Vis/NIR | |
| CN107219184A (en) | Meat identification method and device applied to origin tracing | |
| Mavi et al. | Mango ripeness classification system using hybrid technique | |
| CN113310936A (en) | Rapid identification method for four high-temperature sterilized commercial milks | |
| CN201156043Y (en) | Non-destruction detector for synthetic quality of food | |
| CN107576660A (en) | A kind of double yellow duck egg Automatic Visual Inspection method based on apart from contour | |
| US20210199643A1 (en) | Fluid classification | |
| Thong et al. | Mango sorting mechanical system combines image processing | |
| Valous et al. | Quality evaluation of meat cuts | |
| CN113406013A (en) | Meat product quality detection device and method based on hyperspectral and near-infrared fusion | |
| CN116735610A (en) | Steel pipe surface defect detection method based on machine vision | |
| Xu et al. | An electronic nose system for the monitoring of water cane shoots quality with swarm clustering algorithm |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| EE01 | Entry into force of recordation of patent licensing contract |
Assignee: Jiangsu Yingyang Industrial Co., Ltd. Assignor: Jiangsu University Contract record no.: 2010320000479 Denomination of invention: Agricultural and animal product nondestrctive detection method based on electronic visual sense and smell sense fusion technology and its device Granted publication date: 20060315 License type: Exclusive License Open date: 20041222 Record date: 20100428 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20060315 Termination date: 20150108 |
|
| EXPY | Termination of patent right or utility model |