A Two-Pass Approach to Pattern Classification

Abstract

The work presents an extension of the fuzzy approach to 2-D shape recognition [1] through refinement of initial or coarse classification decisions under a two pass approach. In this approach, an unknown pattern is classified by refining possible classification decisions obtained through coarse classification of the same. To build a fuzzy model of a pattern class horizontal and vertical fuzzy partitions on the sample images of the class are optimized using genetic algorithm. To make coarse classification decisions about an unknown pattern, the fuzzy representation of the pattern is compared with models of all pattern classes through a specially designed similarity measure. Coarse classification decisions are refined in the second pass to obtain the final classification decision of the unknown pattern. To do so, optimized horizontal and vertical fuzzy partitions are again created on certain regions of the image frame, specific to each group of similar type of pattern classes. It is observed through experiments that the technique improves the overall recognition rate from 86.2%, in the first pass, to 90.4% after the second pass, with 500 training samples of handwritten digits.

Microphone array based classification for security monitoring in unstructured environments

“…The idea of coarse-to-grained classification is discussed in [32], and can easily be justified by some informal judgments. In particular, the average frequency spectrum of a weapon presents ample differences from the average spectrum of a vehicle [3], and these differences are easier to learn with respect to the specific differences in spectrum inside each category.…”

A survey on optical character recognition for Bangla and Devanagari scripts

“…Basu et al (2005) have presented an application of the Dempster-Shafer (DS) technique (Ng & Abramson 1990) for combination of classification decisions obtained from two MLP-based classifiers for recognition of handwritten Bangla numerals. Two feature sets, one containing octant-based shadow and centroid features (Basu et al 2004), and another containing longest run features are used to supply digit patterns to MLP classifier. Table 5 gives an overview of the system architectures of the Bangla handwritten numeral recognition schemes.…”

Automation of Indian Postal Documents Written in Bangla and English

“…Many techniques have been proposed for Bangla handwritten numeral recognition [1,2,7,12,13]. Dutta and Chaudhuri [7] proposed a two stage feed-forward NN based scheme for the same.…”

“…The image plane is expanded or trimmed to fit this range. The aspect ratio of the original image is calculated by equation (1). To calculate the mapping function (R 2 ) for the normalized image, we have used square root of the aspect ratio of the original image, given by R 2 = √R 1 .…”