Text Extractions from Natural Scenes by Various Methods

Abstract: Text Extraction plays a major role in finding vital and valuable information. Text extraction involvesdetection, localization, tracking, binarization, extraction, enhancement and recognition of the text fromthe given image. These text characters are difficult to be detected and recognized due to their deviation ofsize, font, style, orientation, alignment, contrast, complex coloured, textured background. Due to rapidgrowth of available multimedia documents and growing requirement for information, identification,indexing and retrieval, many researches have been done on text extraction in images. Several techniqueshave been developed for extracting the text from an image. The proposed methods were based onmorphological operators, wavelet transform, artificial neural network, skeletonization operation, edgedetection algorithm, histogram technique etc. All these techniques have their benefits and restrictions. This article discusses various schemes proposed earlier for extracting the text from an image. This paperalso provides the performance comparison of several existing methods proposed by researchers inextracting the text from an image.

Keywords: Text Extraction, Document Text Images, Caption Text Images, Scene Text, Heterogeneous Images.