Adaptive thresholding technique for document image analysis

Abstract

Binarkation of document images is 1, JntroductionImage binarization or thrcsholding is an important tool in image processing and computer vision, to extract the object pixels in an image from the background pixels. Image binarization is central to many applications including document image analysis (printed characters, logos, graphical content, and musical scores are important as objects), map processing (fines, lzgends and characters necd to be cxtracted), scene processing, quality inspection of materials, cell images, segmentation of various image modalities for nondestructive testing (NDT) applications (ultrasonic images, eddy cumnt images, thermal images, X-ray computed tomography, laser scanning confocal microscopy, extraction of edge field and spatioteinporal segmentation of video images). A number of methods have already been proposed for image binarization but unfortunately, most of them are very much specific for a few applications. Thus, it can be said that a binarization (thresholding) inethod may work well for one application but its performance can be unsatisfactory for another application.Bi-level image is used as a pre-processing unit in several applications. The use of binary images decreases computational load for the overall application. These applications include document analysis, optical character recognition system, scene matching, quality inspection of materials etc. The binarization process computes the threshold value that differentiate object and background pixels. Under varying illumination and noise, the binarization can become a challenging job. A number of factors contribute to complicate the thresholding scheme inchding ambient illumination, variance of gray levels within the object and the background, inadequate contrast, object shape and size non-commensurate with the scene. A wrong selection of threshold value may misinterpret thc background pixel and can classify it as object and vice versa, resulting in overall degradation of system performance. The determination of a threshold itself i s application dependent since one threshold may work with one application and may not work with other one. In document analysis, binarization is sensitive to noisc, surrounding illumination, gray Ievel distribution, local shading effects? inadequate contrast, the presence o f dense nontext components such as photographs, etc. while at the same time, the merges,' fractures and other deformations in the character shapes affects the threshold value io OCR system. On the other hand, the simplification is needed that benefits the overall system processing characteristics (computational load, algorithm complexity, real-time requirements etc). So all these challenges and problems makes binarization a difficult task.There are a number of important performance requirements that need to be considered while binarizing gray Ievel images. With reference to the proposed algorithm, following sections describe all the rehtsd concepts briefly along with the rclatcd work done in the past. The algorithmi...

Benchmarking of Document Image Analysis Tasks for Palm Leaf Manuscripts from Southeast Asia

“…Binarization is widely applied as the first pre-processing step in image document analysis [34]. Binarization is a common starting point for document image analysis and converts gray image values into binary representation for background and foreground, or, more specifically, text and non-text, which is then fed into further document processing tasks such as text line segmentation and optical character recognition.…”

“…Some surveys and comparative studies of the performance of several binarization methods have been reported [35,36]. A binarization method that performs well for one document collection may not necessarily be applied to another document collection with the same performance [34]. For this reason, there is always a need to perform a comprehensive evaluation of the existing binarization methods for a new document collection that has different characteristics, for example the historical archive documents [36].…”

Development of an automated multi-thresholding technique for identification of different materials types and concentration using CT scans

“…This method depends on converting the image into a binary image, which has limited applications. This method can achieve outstanding results in some applications but poor in the others (Rais et al, 2004).…”

“…The global thresholding relies on many ways, such as using the virtual observation of the histogram modes to differentiate between them, or using the trial and error method until finding the suitable threshold according to observer preferences. The local thresholding method uses multiple thresholds for multiple window regions (Gonzales et al, 2006;Rais et al, 2004).…”

Image Segmentation for Uneven Lighting Images using Adaptive Thresholding and Dynamic Window based on Incremental Window Growing Approach

“…The segmented image obtained from thresholding has the advantages of smaller storage space, fast processing speed and ease in manipulation, compared with gray level image which usually contains a large number of gray levels(maximum 256 levels) [3]. Image binarization is central to many applications including document image analysis (printed characters, logos, graphical content), map processing (lines, legends and characters need to be extracted), scene processing, quality inspection of materials, cell images, fingerprint recognition and have a wide range of application in the area of medical image processing [1]. The thresholding techniques can be divided into bilevel and multilevel category.…”