Menu
Home
People
Places
Arts
History
Plants & Animals
Science
Life & Culture
Technology
Reference.org
Bag-of-words model in computer vision
open-in-new
<p>In <a href="/facts/Computer_vision/Tl2Yyk66">computer vision</a>, the bag-of-words (BoW) model, sometimes called bag-of-visual-words model, can be applied to <a href="/facts/Image_classification/Tl2Yyk66">image classification</a> or <a href="/facts/Data_retrieval/Hqe9K9qd">retrieval</a>, by treating <a href="/facts/Image_feature/GdfWRp5U">image features</a> as words. In <a href="/facts/Document_classification/B69lL6BH">document classification</a>, a <a href="/facts/Bag_of_words/TlhIx600">bag of words</a> is a <a href="/facts/Sparse_vector/FFtN0XjR">sparse vector</a> of occurrence counts of words; that is, a sparse <a href="/facts/Histogram/oMmvqGEA">histogram</a> over the vocabulary. In <a href="/facts/Computer_vision/Tl2Yyk66">computer vision</a>, a <i>bag of visual words</i> is a vector of occurrence counts of a vocabulary of local image features. </p>