http://rdf.ncbi.nlm.nih.gov/pubchem/patent/CN-103927555-B
Outgoing Links
| Predicate | Object |
|---|---|
| classificationIPCInventive | http://rdf.ncbi.nlm.nih.gov/pubchem/patentipc/G06K9-62 |
| filingDate | 2014-05-07-04:00^^<http://www.w3.org/2001/XMLSchema#date> |
| grantDate | 2018-01-05-04:00^^<http://www.w3.org/2001/XMLSchema#date> |
| publicationDate | 2018-01-05-04:00^^<http://www.w3.org/2001/XMLSchema#date> |
| publicationNumber | CN-103927555-B |
| titleOfInvention | Static manual alphabet identifying system and method based on Kinect sensor |
| abstract | The present invention relates to computer vision and field of intelligent man-machine interaction, and in particular to a kind of man-machine interactive system and its exchange method based on machine vision.The present invention combines Kinect sensor and proposes that one kind carries out static manual alphabet based on improved SURF algorithm and knows method for distinguishing.The depth image of Kinect sensor collection target area, which carries out hand pixel region division, can overcome the interference that illumination variation, complex background are brought；Improved SURF algorithm is used to extract characteristic point, adaptive radius r is set simultaneously, step-sizing is carried out to SURF characteristic points by 2 comparative feature point number, characteristic point spacing indexs in the neighborhood using r as radius, not only substantially increase discrimination, and ensure that identification be operated in the environmental factors such as the colour of skin, illumination variation, complex background and angle change and dimensional variation in terms of robustness.The problem of in order to overcome SURF feature vector dimensions higher, using SVM " one-to-one " classification, classification based training is carried out to SURF feature descriptors, draws the result of identification. |
| priorityDate | 2014-05-07-04:00^^<http://www.w3.org/2001/XMLSchema#date> |
| type | http://data.epo.org/linked-data/def/patent/Publication |
Incoming Links
Total number of triples: 13.