Variations in Shape Context Descriptor: A Survey

Bhuptani, Neha; Talati, Bjial

doi:10.5120/15775-4541

Cited by 2 publications

(6 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Hence, shape descriptors are important. Efficient shape representation must fulfil some requisites like translation invariance, rotation invariance, scale invariance, and indentifiability, noise resistant and reliable [1]. The shape descriptor is evaluated by the requisites that it fulfils.…”

Section: Fig 1 General Workflow Of Cbirmentioning

confidence: 99%

“…Due to its robustness to rotation and scaling and its usefulness in object identification, shape context (SC) is employed as the global shape feature [1]. It describes the coarse distribution of the rest of the shape with respect to a given point on the shape.In this paper, shape context is explained in detail and improvement is made in it to reduce the computational cost by maintaining its shape retrieval effectiveness.…”

Section: Fig 1 General Workflow Of Cbirmentioning

confidence: 99%

“…As discussed in [1], Fig 2 [6] gives the basic idea of shape contexts. The random sample points of the two shapes are taken and shown in the Fig.…”

Section: Shape Context and Related Workmentioning

confidence: 99%

“…As a result it could not be applied to huge database. So many variations were studied in survey done in [1].A survey of different variations of shape context could be seen in it to overcome different problems of shape context.…”

Section: Fig 3 Proposed Workmentioning

confidence: 99%

“…Now this shape context of centroid of query image is matched with the shape context of the centroid of the images in the database. The distance between two centroidal shape contexts is calculated using chi square distance which is defined as given in equation (1).…”

Section: Preliminary Filtering Stagementioning

confidence: 99%

See 4 more Smart Citations

An Efficient Image Retrieval Technique using Shape Context Feature

Bhuptani¹,

Talati²

2014

IJCA

Self Cite

View full text Add to dashboard Cite

Increasing demands for image retrieval in field such as health informatics, biometrics, and crime prevention has forced developers to explore ways to manage and retrieve images more efficiently. An automated way to retrieve images based on the content or features of the images itself is provided by CBIR. Shape is one of key visual features used by human for distinguishing visual data along with other features of color and texture. Therefore, this work investigates one of the shape representation method shape context which helps in efficient CBIR.The main aim of this research is to look for and develop promising shape descriptor(s) which was found to be Shape Context for image retrieval and also to improve efficiency. The Shape context descriptor is contour based which focuses on irregular shapes. The time consuming step of shape context matching is reduced in this work up to approx. 48% on an average. The proposed work reduces the computational complexity maintaining its overall accuracy.

show abstract

Section: Fig 1 General Workflow Of Cbirmentioning

confidence: 99%

Section: Fig 1 General Workflow Of Cbirmentioning

confidence: 99%

“…As discussed in [1], Fig 2 [6] gives the basic idea of shape contexts. The random sample points of the two shapes are taken and shown in the Fig.…”

Section: Shape Context and Related Workmentioning

confidence: 99%

Section: Fig 3 Proposed Workmentioning

confidence: 99%

Section: Preliminary Filtering Stagementioning

confidence: 99%

See 3 more Smart Citations

An Efficient Image Retrieval Technique using Shape Context Feature

Bhuptani¹,

Talati²

2014

IJCA

Self Cite

View full text Add to dashboard Cite

show abstract

Pengenalan Tulisan Pada Iklan Pinggir Jalan yang Melengkung Menggunakan Shape Context

Setyati

Sugiarto

2021

INSYST

View full text Add to dashboard Cite

Membaca sebuah tulisan yang sama di bidang melengkung berbeda dengan di bidang datar, karena tulisan pada bidang melengkung bergantung pada permukaan bidang lengkungnya. Pada saat ini, banyak sekali tulisan pada iklan pinggir jalan yang ditempel pada bidang melengkung di sepanjang jalan. Tulisan yang digunakan berupa huruf dan angka, dengan berbagai macam background, bentuk dan warna yang diambil di pinggir jalan dengan menggunakan Farey Shape Context. Fitur Farey ini bergantung pada DSS (Digital Straight Line Segment) endpoint dan menggunakan pecahan Augmented Farey sequence. DSS endpoint ini dijadikan sebagai titik fitur atau feature point untuk menemukan shape context dari citra. DSS endpoint tersebut digunakan sebagai acuan bounding box yang akan digunakan sebagai object boundary yang dimana setiap sudutnya merupakan reference point. Untuk melakukan Binning Farey Rank, Augmented Farey Table (AFT) harus dibentuk terlebih dahulu berdasarkan Augmented Farey Sequence yang merupakan pengembangan dari Farey Sequence. Farey Sequence hanya meliputi pecahan dengan pembilang dan penyebut yang positif, sedangkan Augmented Farey Sequence meliputi pecahan dengan pembilang dan penyebut positif serta negatif. Pada penelitian ini digunakan 500 data iklan di pinggir jalan yang melengkung, dimana 70% digunakan sebagai data sample. Dari 70% data sample tersebut didapatkan ribuan karakter berupa huruf dan angka yang dijadikan data sample. Berdasarkan hasil uji coba penelitian yang dilakukan pada 500 Gambar dimana 30% sebagai data testing, maka hasil Farey Shape Context untuk mengenali tulisan berupa huruf dan angka pada iklan pinggir jalan yang melengkung mencapai akurasi benar 74.94% dan salah 25.06%.

show abstract

Variations in Shape Context Descriptor: A Survey

Cited by 2 publications

References 11 publications

An Efficient Image Retrieval Technique using Shape Context Feature

An Efficient Image Retrieval Technique using Shape Context Feature

Pengenalan Tulisan Pada Iklan Pinggir Jalan yang Melengkung Menggunakan Shape Context

Contact Info

Product

Resources

About