International Scholarly Research Notices / 2012 / Article / Fig 5

Research Article

An Effective Color Addition to Feature Detection and Description for Book Spine Image Matching

Figure 5

The source image in each channel is independently convolved with a Gaussian kernel at differing values of 𝜎 , and then adjacent 𝜎 images are differenced to produce a Difference of Gaussians. The π‘Œ channel DoG is shown in (b), the 𝐢 π‘Ÿ in (c), and the 𝐢 𝑏 in (d). Different strong edges are visible in each channel. A grayscale feature detector would only find those visible in the π‘Œ channel and miss the very obvious features in the chrominance channels.
945973.fig.005a
(a) Original book spine image
945973.fig.005b
(b) DoG image in π‘Œ channel
945973.fig.005c
(c) DoG image in 𝐢 π‘Ÿ channel
945973.fig.005d
(d) DoG image in 𝐢 𝑏 channel

Article of the Year Award: Outstanding research contributions of 2020, as selected by our Chief Editors. Read the winning articles.