Blind document image quality prediction based on modification of quality aware clustering method integrating a patch selection strategy

Abstract

The quality of document images has direct impacts on the performance of document image processing systems. Document Image Quality Assessment (DIQA) is, therefore, of fundamental importance to a numerous document processing applications. As manual quality assessment is almost impossible for a huge volume of document images generated in day-today life, it is critical to develop intelligent machine operated methods to estimate the quality of document images. In this paper, a blind document image quality assessment method is proposed to deal with the problem of DIQA in real scenarios, as reference images are not always available. To estimate the quality of a document image, the document is rst sampled into a set of patches. The extracted patches are then ltered out based on their level of foreground information using a patch selection strategy. For every selected patch, a cluster assignment is then performed to obtain its quality from a quality aware bag of visual words constructed using k-means clustering. An average pooling is nally employed to estimate the quality of the input document image. To evaluate the proposed method, a dataset composed of document images and three scene image datasets were considered for experimentation. The results obtained from the proposed method demonstrate the eectiveness of the proposed DIQA method. These achievements in applied computational intelligence, expert and decision support systems make a good foundation for creating practical tools to automate document image forgery detection, and archiving process.

No Reference Quality Assessment for Screen Content Images Using Stacked Autoencoders in Pictorial and Textual Regions

“…Alaei et al [34] proposed a FR DIQA method according to the hypothesis that the human perception of the document can be influenced by the foreground information. Subsequently, Alaei et al [35] presented a NR DIQA metric via modification of the quality-aware clustering method which integrates a patch selection strategy. Ye et al [36] designed an unsupervised feature learning method by learning features for predicting the accuracy of OCR.…”

Handwritten Arabic Bills Reader and Recognizer

“…The training process incorporates the use of linear regression. Similarly, Alaei et al [20] utilized an unsupervised approach, trained on the ITESOFT dataset, in which the image was partitioned into patches to generate a Bag of Words (BoW) representation. In the testing phase, features extracted from these patches are allocated to clusters and subjected to average pooling.…”

Document Image Forgery and Detection Methods Using Image Processing Techniques – A Review

“…Alireza et al [29] have described how to evaluate the quality of a blind document to solve DIQA issues in realworld scenarios, as reference images are not always available. It is first sampled into a series of patches to measure the quality of the document.…”