http://rdf.ncbi.nlm.nih.gov/pubchem/patent/CN-105005772-B
Outgoing Links
Predicate | Object |
---|---|
classificationCPCInventive | http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06V20-41 http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06V20-46 |
classificationIPCInventive | http://rdf.ncbi.nlm.nih.gov/pubchem/patentipc/G06K9-00 |
filingDate | 2015-07-20-04:00^^<http://www.w3.org/2001/XMLSchema#date> |
grantDate | 2018-06-12-04:00^^<http://www.w3.org/2001/XMLSchema#date> |
publicationDate | 2018-06-12-04:00^^<http://www.w3.org/2001/XMLSchema#date> |
publicationNumber | CN-105005772-B |
titleOfInvention | A kind of video scene detection method |
abstract | The invention discloses a kind of video scene detection method, this method, for being manually detected to video data, identifies the scene in video by computer generation;Detection method includes off-line training discrimination model process and video scene detection process.Off-line training discrimination model process for video extraction feature each in training video sample set, is extracted including semantic and space-time characteristic;Classification mark is carried out to feature vector, has obtained one group of sample set;Training is iterated to sample set using Multiple Kernel Learning frame, obtains off-line training model;Video scene detection process access monitor video source;It carries out video sampling and obtains a short-sighted frequency;To short-sighted frequency extraction feature;It is loaded into off-line training model to be detected feature, obtains testing result.The present invention, for the scene in manual identified video, can be improved detection efficiency, reduce cost, facility is also provided to data storage and search by computer generation. |
isCitedBy | http://rdf.ncbi.nlm.nih.gov/pubchem/patent/CN-110166744-A |
priorityDate | 2015-07-20-04:00^^<http://www.w3.org/2001/XMLSchema#date> |
type | http://data.epo.org/linked-data/def/patent/Publication |
Incoming Links
Predicate | Subject |
---|---|
isDiscussedBy | http://rdf.ncbi.nlm.nih.gov/pubchem/compound/CID86122 http://rdf.ncbi.nlm.nih.gov/pubchem/substance/SID415781046 |
Total number of triples: 14.