http://rdf.ncbi.nlm.nih.gov/pubchem/patent/CN-110163836-B
Outgoing Links
Predicate | Object |
---|---|
classificationCPCAdditional | http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06T2207-20084 http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06T2207-20081 |
classificationCPCInventive | http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06T7-001 |
classificationIPCInventive | http://rdf.ncbi.nlm.nih.gov/pubchem/patentipc/G06T7-00 |
filingDate | 2018-11-14-04:00^^<http://www.w3.org/2001/XMLSchema#date> |
grantDate | 2021-04-06-04:00^^<http://www.w3.org/2001/XMLSchema#date> |
publicationDate | 2021-04-06-04:00^^<http://www.w3.org/2001/XMLSchema#date> |
publicationNumber | CN-110163836-B |
titleOfInvention | Excavator detection method used under high-altitude inspection based on deep learning |
abstract | The method for detecting the excavator under the high-altitude inspection based on the deep learning comprises the steps of manufacturing an excavator data set under a high-altitude lower visual angle, and adopting a proper data augmentation strategy according to the visual angle and the environmental characteristics. A detection frame is constructed by using a neural network, and a backbone network adopts a full convolution neural network with a 43-layer convolution structure; the detection frame comprises a bottom-up path, a top-down path and a side path; clustering six anchor frames by a k-means method, distributing large-size anchor frames to 38-38 feature map operation and matching, and distributing small-size anchor frames to 76-76 feature map operation and matching; the detector is trained. The invention has the advantages of high detection accuracy and small operation burden. |
priorityDate | 2018-11-14-04:00^^<http://www.w3.org/2001/XMLSchema#date> |
type | http://data.epo.org/linked-data/def/patent/Publication |
Incoming Links
Total number of triples: 15.