Javascript must be enabled for the correct page display

Local quantisation based on Robust Automatic Threshold Selection

Broek, R. van den (2004) Local quantisation based on Robust Automatic Threshold Selection. Master's Thesis / Essay, Computing Science.

Infor_Ma_2004_RvandenBroek.CV.pdf - Published Version

Download (1MB) | Preview


Binarisation is a common image operation, which separates an image into two classes, and there are many algorithms for this operation. However, there are fewer algorithms for the generalisation of binarisation to multiple classes, quantisation. Robust Automatic Threshold Selection (RATS) is a well-known technique for binarisation based on a simple image statistic, due to Kittler eta!. [4J. However, as opposed to some other methods, such as Otsu's algorithm [7], RATS does not readily extend to quantisation. In this paper, we will propose a method to perform quantisation using an adaptation of RATS. We will prove that this method is correct with respect to step edges, and use it as a basis for a local quantisation algorithm. We will experiment with the local version of the algorithm, which segments an image into a number of objects, to find good values for the algorithm's parameters. Finally, we will show some results of applying the local algorithm to a number of real-life examples.

Item Type: Thesis (Master's Thesis / Essay)
Degree programme: Computing Science
Thesis type: Master's Thesis / Essay
Language: English
Date Deposited: 15 Feb 2018 07:30
Last Modified: 15 Feb 2018 07:30

Actions (login required)

View Item View Item