http://rdf.ncbi.nlm.nih.gov/pubchem/patent/CN-107101972-B
Outgoing Links
Predicate | Object |
---|---|
classificationCPCInventive | http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G01N1-286 http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G01N21-359 |
classificationIPCInventive | http://rdf.ncbi.nlm.nih.gov/pubchem/patentipc/G01N21-359 http://rdf.ncbi.nlm.nih.gov/pubchem/patentipc/G01N1-28 |
filingDate | 2017-05-24-04:00^^<http://www.w3.org/2001/XMLSchema#date> |
grantDate | 2019-10-15-04:00^^<http://www.w3.org/2001/XMLSchema#date> |
publicationDate | 2019-10-15-04:00^^<http://www.w3.org/2001/XMLSchema#date> |
publicationNumber | CN-107101972-B |
titleOfInvention | A kind of near infrared spectrum quickly detects radix tetrastigme place of production method |
abstract | The invention discloses a kind of methods that near infrared spectrum quickly identifies the radix tetrastigme place of production.Since radix tetrastigme near infrared spectrum data is not also perfect, in its place of production discriminating, reply identification algorithm is put forward higher requirements, i.e., while realizing three leaf place of production discriminating, also wants the radix tetrastigme that can efficiently identify out the other or unknown new place of production.For this problem, the characteristics of herein in conjunction with radix tetrastigme near infrared spectrum data, three aspects is done altogether to algorithm and have improved: 1. from the probability density of the angle estimation sample of distance;2. the computation bandwidth parameter in a manner of training sample confidence level;3. proposing a kind of recognition methods of probability density function based on training set sample in the identification in the unknown new place of production.The present invention proposes a kind of new characteristic wavelength selection method, easy to operate, versatile, can effectively improve quantitative model, robustness and accuracy. |
priorityDate | 2017-05-24-04:00^^<http://www.w3.org/2001/XMLSchema#date> |
type | http://data.epo.org/linked-data/def/patent/Publication |
Incoming Links
Total number of triples: 16.