http://rdf.ncbi.nlm.nih.gov/pubchem/patent/CN-105405107-B
Outgoing Links
| Predicate | Object |
|---|---|
| classificationCPCAdditional | http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06T2207-20081 http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06T2207-20024 |
| classificationCPCInventive | http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06T5-003 |
| classificationIPCInventive | http://rdf.ncbi.nlm.nih.gov/pubchem/patentipc/G06T5-00 |
| filingDate | 2015-10-23-04:00^^<http://www.w3.org/2001/XMLSchema#date> |
| grantDate | 2018-05-08-04:00^^<http://www.w3.org/2001/XMLSchema#date> |
| publicationDate | 2018-05-08-04:00^^<http://www.w3.org/2001/XMLSchema#date> |
| publicationNumber | CN-105405107-B |
| titleOfInvention | A kind of image super-resolution rebuilding method |
| abstract | The invention discloses a kind of image super-resolution rebuilding method, including：To input picture Y L Gassian low-pass filter and bicubic up-sampling are carried out respectively, obtain Gassian low-pass filter image X L With bicubic up-sampling image X H ；According to input picture Y L With Gassian low-pass filter image X L Image is built to training set Then method is searched using nearest-neighbor to training set D according to image and obtains image X H Training sample to set；Using multitask Gaussian process regression model to image X H Training sample to set be described, then using gradient descent method carry out parameter training, obtain image X H Multitask Gauss model parameter and multitask Gauss model output Y corresponding to sample H ；Y is exported to the multitask Gauss model obtained H Final super-resolution image is obtained using back projection method.The present invention has the advantages that speed is fast and effect is good, can be widely applied to image processing field. |
| priorityDate | 2015-10-23-04:00^^<http://www.w3.org/2001/XMLSchema#date> |
| type | http://data.epo.org/linked-data/def/patent/Publication |
Incoming Links
Total number of triples: 18.