A Simple Shape Descriptor Merging Arithmetical Wrap Around Technique with Absolute Localized Pixel Differences
- PDF / 1,885,938 Bytes
- 17 Pages / 439.37 x 666.142 pts Page_size
- 13 Downloads / 166 Views
A Simple Shape Descriptor Merging Arithmetical Wrap Around Technique with Absolute Localized Pixel Differences Kethepalli Mallikarjuna1 · Bepar Abdul Raheem2 · Govindaraj Pathanadka3 · Sudhakar Mogappair Suriyakumar4 Accepted: 11 November 2020 © Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract The quest for computationally simple, highly accurate and precise shape descriptors supporting retrieval continues to be an active research area in computer vision. In this paper, a simple feature descriptor is realized by blending Modulo Arithmetic (MA) with Local Absolute Pixel Differences (LAPD) labelled as MA-LAPD. MA initially refines edges of images through modulo normalization and later operated by LAPD to capture local texture patterns. Subsequently, LAPD encodes the local intensity transitions in eight directions with regard to center pixel. The two prominent directional indices are converted into unique decimal codes that represent each pixel position, thus, transforming each image into a collection of LAPD codes. The ensuing LAPD image is then fabricated into histograms for characterizing the distribution of local features, used for matching and retrieval. Quantitative and qualitative investigations on Kimia’s 99, MPEG-7 Part-B and Tari-1000 datasets reveal consistent Bull’s Eye Retrieval (BER) scores above 91%. Also, relative analysis exposes the superiority of MA-LAPD with its predecessors in majority of the datasets. Keywords Bull’s eye retrieval · Image histogram · Local absolute pixel differences · Modulo arithmetic · Shape retrieval
* Kethepalli Mallikarjuna [email protected] Bepar Abdul Raheem [email protected] Govindaraj Pathanadka [email protected] Sudhakar Mogappair Suriyakumar [email protected] 1
RGM College of Engineering and Technology, Nandyala, India
2
Annamacharya Institute of Technology and Sciences, Rajampet, India
3
AJ Institute of Engineering and Technology, Manglore, Mangalore, India
4
Vellore Institute of Technology, Vellore, India
13
Vol.:(0123456789)
K. Mallikarjuna et al.
1 Introduction Shape recognition and classification fields of computer vision highly rely on shape for feature characterization and description. Also, the size of the resulting descriptor can be reduced by defining it in terms of the shape feature. Moreover, shape descriptors abiding invariance are deemed fit for applications such as image and video retrieval, image segmentation and recognition [1]. The inherent characteristic of human visual system interprets images as geometrical shapes by visual decomposition. Further, shapes offer the ability to discern the variations within intra and interclasses. Prevailing shape matching and retrieval schemes offered escalated results over the years by covering affine transformations such as translation, scaling and rotation. Whilst, achieving such quality in shape matching is challenging and cumbersome especially in cases of non-linear deformations. Moreover, the complexity increases with incr
Data Loading...
