Matches in SemOpenAlex for { <https://semopenalex.org/work/W2766021910> ?p ?o ?g. }
Showing items 1 to 46 of
46
with 100 items per page.
- W2766021910 abstract "Human Activity detection is an imperative area of research in computer vision. This paper focuses on activity recognition by construction personnel at the construction sites. The method uses bag of features (BOF) approach to detect an activity. Here we have considered five types of activities done at construction sites namely ladder climbing, brick laying, carpentry work, painting and plastering work. Still images of activities at construction sites are obtained from Internet and few were added from personal digital collection to form a new construction workers database, comprising a total of 389 images. 30% of the images from the dataset are used for training, 40% for validation and 30% for testing. Bag of features classifier is applied to the training set to form 500 visual words. The test images are classified as one of the five categories with an accuracy of 95%. The proposed method is also tested on KTH and willows action dataset. The accuracy for KTH dataset is 96% and for Willows action dataset is 95%. All algorithms are tested using MATLAB 2012b software." @default.
- W2766021910 created "2017-11-10" @default.
- W2766021910 creator A5085977948 @default.
- W2766021910 date "2017-03-01" @default.
- W2766021910 modified "2023-09-24" @default.
- W2766021910 title "Construction Workers Activity Detection Using BOF" @default.
- W2766021910 cites W1589091100 @default.
- W2766021910 cites W1807871095 @default.
- W2766021910 cites W1966941463 @default.
- W2766021910 cites W1993991024 @default.
- W2766021910 cites W2008748781 @default.
- W2766021910 cites W2022329870 @default.
- W2766021910 cites W2100916003 @default.
- W2766021910 cites W2119605622 @default.
- W2766021910 cites W2158169396 @default.
- W2766021910 cites W2381252496 @default.
- W2766021910 cites W2533503513 @default.
- W2766021910 cites W2533739470 @default.
- W2766021910 doi "https://doi.org/10.1109/icraect.2017.54" @default.
- W2766021910 hasPublicationYear "2017" @default.
- W2766021910 type Work @default.
- W2766021910 sameAs 2766021910 @default.
- W2766021910 citedByCount "2" @default.
- W2766021910 countsByYear W27660219102020 @default.
- W2766021910 countsByYear W27660219102022 @default.
- W2766021910 crossrefType "proceedings-article" @default.
- W2766021910 hasAuthorship W2766021910A5085977948 @default.
- W2766021910 hasConcept C41008148 @default.
- W2766021910 hasConceptScore W2766021910C41008148 @default.
- W2766021910 hasLocation W27660219101 @default.
- W2766021910 hasOpenAccess W2766021910 @default.
- W2766021910 hasPrimaryLocation W27660219101 @default.
- W2766021910 hasRelatedWork W1596801655 @default.
- W2766021910 hasRelatedWork W2049775471 @default.
- W2766021910 hasRelatedWork W2350741829 @default.
- W2766021910 hasRelatedWork W2358668433 @default.
- W2766021910 hasRelatedWork W2376932109 @default.
- W2766021910 hasRelatedWork W2382290278 @default.
- W2766021910 hasRelatedWork W2390279801 @default.
- W2766021910 hasRelatedWork W2748952813 @default.
- W2766021910 hasRelatedWork W2899084033 @default.
- W2766021910 hasRelatedWork W2530322880 @default.
- W2766021910 isParatext "false" @default.
- W2766021910 isRetracted "false" @default.
- W2766021910 magId "2766021910" @default.
- W2766021910 workType "article" @default.