Skew detection and correction document

                                                               

Skew detection and correction is very important task before pre-processing of an image and it is a major problem in scanned documents, if itnot detected correctly it will lead wrong result in future during image analysis. During the scanning of the document, skew is being often introduced in the

document image. To measure the processed time and speed taken by skew detection algorithm, the Fast Fourier Transform (FFT) technique is applied as it is fast approach for finding the angle of skewed document. This technique is used by firstly applying DCT compression and thresholding on image to reduce timing computation and after that Fourier spectrum is obtained. Further this spectrum is divided into four quadrants and detected skewed angle of each quadrant is measured. And finally Input image is rotated by using bilinear interpolation method

Skew angle estimation and correction of a scanned document is very important task for document analysis. During digitization of documents, it often happens that the document is not aligned

correctly and it may lead to skewed image. Therefore, due to skew it can cause further performance degradation of segmentation and recognition stage of any text processing system [1]. Skew in scanned document can be of two types and

shown in figure 1 [2].

1. Clock -wise skew (Positive skew)
2. Anti-clock wise skew (Negative skew)

In 2012 Kumar et al. has introduced a new method which reduces the time complexity without compromising with the accuracy of Hough transform. The main advantage of Hough transform [1] is due to its better accuracy and simplicity. In 2002 Lowether et al. [6] has presented a new Averaged Block Directional Spectrum (ABDS) technique for determining the skew angle of digitized documents. It is based on calculating the average 2D Fourier Transform of blocks in a document image and using the Radon transform to find the peak in the directional spectrum. In 2007 Aradhya et al. [7] has proposed a novel skew detection method for binary document images. This method considered some selected characters of the text which may be subjected to thinning and after applying Hough transform skew angle of the documents are estimated. On the other hand, several experiments have been conducted on various types of documents such as English documents, Journals, Text-Books, different languages, documents with different fonts and resolutions etc. to reveal the robustness of the proposed method. The experimental results reveal that the proposed method is accurate in comparison to the results of well-known existing methods

Conclusion

In this paper, a new technique is proposed for determining the skew angle of digitized documents. The results obtained are highly accurate and less time consuming with better speed as compare to the other existing methods. The accuracy achieved a skew angle determination within the range of ±45 degree of true skew angle. The advantage of the technique over most other techniques is the ease of detecting skew over ±45 degree skewed angle. This technique calculates the processed time of skewed angle of different handwritten English documents, Journals, Text-Books, different languages,

documents with different fonts and printed documents. Accuracy is also near to 95% and time for calculating the skewed angle is less than one second (average) for all scanned images.