Automatic Discrimination between Printed and Handwritten Text in Documents

Abstract

Recognition techniques for printed and handwritten text in scanned documents are significantly different. In this paper we address the problem of identifying each type. We can list at least four steps: digitalization, preprocessing, feature extraction and decision or classification. A new aspect of our approach is the use of data mining techniques on the decision step. A new set of features extracted of each word is proposed as well. Classification rules are mining and used to discern printed text from handwritten. The proposed system was tested in two public image databases. All possible measures of efficiency were computed achieving on every occasion quantities above 80%.

Handwritten and Machine-Printed Text Discrimination Using a Template Matching Approach

“…The noise in the data resulted in binarisation failure and missing data in Figure 7-d, while the overlapping line created an occluded segment in Figure 7-c. On the other hand, although the handwritten segments shown in Figure 7-e to -h have close regularities with machine-printed text, they are correctly classified as handwritten since none of the gallery characters can produce a matching score higher than the threshold T . Because of the high similarity with the machine-printed sample, the geometric features, such as area and rectangularity used in [6], [7], [10], [12], fail to create separable classes and the samples will wrongly be classified as machine-printed. Table 1 shows the comparison of our approach with the results provided by Zagoris et al [3], [4], in which 15% of the samples in the PRImA-NHM dataset are utilised for train and the remaining 85% for the test phase.…”

“…The features consist of geometrical, statistical moments, and contours histograms. In [6], eleven features mainly based on the ratio of the statistical and geometrical features of the segmented words to the geometrical size of their bounding boxes (width or height) are extracted from the regions within the bounding boxes.…”

“…Quantitative evaluations have been conducted over the publicly available Pattern Recognition & Image Analysis Research LabNatural History Museum (PRImA-NHM) dataset [3], giving compelling results as compared with the state-of-the-art. The main contributions of the proposed HMC algorithm are as follows: 1) As opposed to [4], [5], [6], [7], [8], [9], [10], the proposed algorithm is training-free as does not rely on a trainable classifier. This provides the capability to update the gallery samples, without re-designing or modifying the decision making step.…”

“…The application of the proposed gallery generation can be extended to perform HMC for other languages. 2) Unlike the geometry-based feature extraction algorithms used by [6], [7], [10], [11], [12], [13], which are sensitive to missing data, occlusion or cluttered text, the proposed approach is robust against occlusions and noise over the scanned text.…”

Text Line Segmentation With Water Flow Algorithm Based on Power Function

“…These methods primarily use the horizontal projection profile, which is obtained by summing pixel values along the horizontal axis. It is achieved by finding its maximum and minimum [5][6]. Each local maximum represents the text line, while local minimum means interline spacing.…”