Matches in SemOpenAlex for { <https://semopenalex.org/work/W4225103640> ?p ?o ?g. }
Showing items 1 to 73 of
73
with 100 items per page.
- W4225103640 abstract "Defect detection of high temperature ceramic chips is an important part of fuel cell production and quality management. Therefore, considering problems that ceramic chip itself is complex and there is not sufficient existing defect detection technology, a defect detection method of ceramic chip based on improved Faster R-CNN is proposed in this paper. After the data is first enhanced, the feature pyramid network structure (FPN) will be introduced into the backbone network, thus generating feature maps suitable for multi-scale target detection to strengthen the feature extraction ability of small target defects. Then, the ROI Align algorithm is used to replace the ROI Pooling algorithm, the maximum pooling operation is performed to obtain a fixed-dimensional ROI output so that more accurate defect location information can be gained. Experimental results show that compared to SSD, Faster R-CNN deep learning model and YOLOv5 model based on VGG-16, ResNet-50, ResNet-101 feature extraction network, the improved model converges fast and detects better in small target defects. In addition, compared with the original model based on VGG-16, its detection accuracy, recall rate and mAP value are respectively increased by 4.46 %, 2.46 % and 6.78 %, meeting the requirements of detection tasks." @default.
- W4225103640 created "2022-04-30" @default.
- W4225103640 creator A5007536681 @default.
- W4225103640 creator A5010055509 @default.
- W4225103640 creator A5024252743 @default.
- W4225103640 creator A5026734277 @default.
- W4225103640 creator A5035113720 @default.
- W4225103640 creator A5040519711 @default.
- W4225103640 creator A5090535728 @default.
- W4225103640 date "2022-02-01" @default.
- W4225103640 modified "2023-09-23" @default.
- W4225103640 title "Research on Defect Recognition of Ceramic Chips for High Temperature Fuel Cells Based on Improved Faster R-CNN" @default.
- W4225103640 doi "https://doi.org/10.1109/mlke55170.2022.00034" @default.
- W4225103640 hasPublicationYear "2022" @default.
- W4225103640 type Work @default.
- W4225103640 citedByCount "0" @default.
- W4225103640 crossrefType "proceedings-article" @default.
- W4225103640 hasAuthorship W4225103640A5007536681 @default.
- W4225103640 hasAuthorship W4225103640A5010055509 @default.
- W4225103640 hasAuthorship W4225103640A5024252743 @default.
- W4225103640 hasAuthorship W4225103640A5026734277 @default.
- W4225103640 hasAuthorship W4225103640A5035113720 @default.
- W4225103640 hasAuthorship W4225103640A5040519711 @default.
- W4225103640 hasAuthorship W4225103640A5090535728 @default.
- W4225103640 hasConcept C134132462 @default.
- W4225103640 hasConcept C138885662 @default.
- W4225103640 hasConcept C142575187 @default.
- W4225103640 hasConcept C153180895 @default.
- W4225103640 hasConcept C154945302 @default.
- W4225103640 hasConcept C159985019 @default.
- W4225103640 hasConcept C165005293 @default.
- W4225103640 hasConcept C192562407 @default.
- W4225103640 hasConcept C2524010 @default.
- W4225103640 hasConcept C2776401178 @default.
- W4225103640 hasConcept C33923547 @default.
- W4225103640 hasConcept C41008148 @default.
- W4225103640 hasConcept C41895202 @default.
- W4225103640 hasConcept C52622490 @default.
- W4225103640 hasConcept C70437156 @default.
- W4225103640 hasConcept C76155785 @default.
- W4225103640 hasConceptScore W4225103640C134132462 @default.
- W4225103640 hasConceptScore W4225103640C138885662 @default.
- W4225103640 hasConceptScore W4225103640C142575187 @default.
- W4225103640 hasConceptScore W4225103640C153180895 @default.
- W4225103640 hasConceptScore W4225103640C154945302 @default.
- W4225103640 hasConceptScore W4225103640C159985019 @default.
- W4225103640 hasConceptScore W4225103640C165005293 @default.
- W4225103640 hasConceptScore W4225103640C192562407 @default.
- W4225103640 hasConceptScore W4225103640C2524010 @default.
- W4225103640 hasConceptScore W4225103640C2776401178 @default.
- W4225103640 hasConceptScore W4225103640C33923547 @default.
- W4225103640 hasConceptScore W4225103640C41008148 @default.
- W4225103640 hasConceptScore W4225103640C41895202 @default.
- W4225103640 hasConceptScore W4225103640C52622490 @default.
- W4225103640 hasConceptScore W4225103640C70437156 @default.
- W4225103640 hasConceptScore W4225103640C76155785 @default.
- W4225103640 hasFunder F4320337504 @default.
- W4225103640 hasLocation W42251036401 @default.
- W4225103640 hasOpenAccess W4225103640 @default.
- W4225103640 hasPrimaryLocation W42251036401 @default.
- W4225103640 hasRelatedWork W2016461833 @default.
- W4225103640 hasRelatedWork W2291847203 @default.
- W4225103640 hasRelatedWork W2546942002 @default.
- W4225103640 hasRelatedWork W2944724518 @default.
- W4225103640 hasRelatedWork W2951391129 @default.
- W4225103640 hasRelatedWork W2969680539 @default.
- W4225103640 hasRelatedWork W3197541072 @default.
- W4225103640 hasRelatedWork W4225258897 @default.
- W4225103640 hasRelatedWork W4320802741 @default.
- W4225103640 hasRelatedWork W3127217315 @default.
- W4225103640 isParatext "false" @default.
- W4225103640 isRetracted "false" @default.
- W4225103640 workType "article" @default.