William B. Lund

Improving optical character recognition through efficient multiple system alignment

Individual optical character recognition (OCR) engines vary in the types of errors they commit in recognizing text, particularly poor quality text. By aligning the output of multiple OCR engines and taking advantage of the differences between them, the error rate based on the aligned lattice of recognized words is significantly lower than the individual OCR word error rates. This lattice error rate constitutes a lower bound among aligned alternatives from the OCR output. Results from a collection of poor quality mid-twentieth century typewritten documents demonstrate an average reduction of 55.0% in the error rate of the lattice of alternatives and a realized word error rate (WER) reduction of 35.8% in a dictionary-based selection process. As an important precursor, an innovative admissible heuristic for the A* algorithm is developed, which results in a significant reduction in state space exploration to identify all optimal alignments of the OCR text output, a necessary step toward the construction of the word hypothesis lattice. On average 0.0079% of the state space is explored to identify all optimal alignments of the documents.

Combining multiple thresholding binarization values to improve OCR output

For noisy, historical documents, a high optical character recognition (OCR) word error rate (WER) can render the OCR text unusable. Since image binarization is often the method used to identify foreground pixels, a body of research seeks to improve image-wide binarization directly. Instead of relying on any one imperfect binarization technique, our method incorporates information from multiple simple thresholding binarizations of the same image to improve text output. Using a new corpus of 19th century newspaper grayscale images for which the text transcription is known, we observe WERs of 13.8% and higher using current binarization techniques and a state-of-the-art OCR engine. Our novel approach combines the OCR outputs from multiple thresholded images by aligning the text output and producing a lattice of word alternatives from which a lattice word error rate (LWER) is calculated. Our results show a LWER of 7.6% when aligning two threshold images and a LWER of 6.8% when aligning five. From the word lattice we commit to one hypothesis by applying the methods of Lund et al. (2011) achieving an improvement over the original OCR output and a 8.41% WER result on this data set.

Error Correction with In-domain Training across Multiple OCR System Outputs

Optical character recognition (OCR) systems differ in the types of errors they make, particularly in recognizing characters from degraded or poor quality documents. The problem is how to correct these OCR errors, which is the first step toward more effective use of the documents in digital libraries. This paper demonstrates the degree to which the word error rate (WER) can be reduced using a decision list on a combination of textual features across the aligned output of multiple OCR engines where in-domain training data is available. This research was performed on a data set for which the mean WER across the three OCR engines employed is 33.5%, and the lattice word error rate is 13.0%. Our correction method leads to a 52.2% relative decrease in the mean WER and a 19.5% relative improvement over the best single OCR engine, as well as an improvement over our previous work. Further, our method yields instances where the document WER approaches and for five documents matches the lattice word error rate, which is a theoretical lower bound given the evidence found in the OCR.

How well does multiple OCR error correction generalize

As the digitization of historical documents, such as newspapers, becomes more common, the need of the archive patron for accurate digital text from those documents increases. Building on our earlier work, the contributions of this paper are: 1. in demonstrating the applicability of novel methods for correcting optical character recognition (OCR) on disparate data sets, including a new synthetic training set, 2. enhancing the correction algorithm with novel features, and 3. assessing the data requirements of the correction learning method. First, we correct errors using conditional random fields (CRF) trained on synthetic training data sets in order to demonstrate the applicability of the methodology to unrelated test sets. Second, we show the strength of lexical features from the training sets on two unrelated test sets, yielding a relative reduction in word error rate on the test sets of 6.52%. New features capture the recurrence of hypothesis tokens and yield an additional relative reduction in WER of 2.30%. Further, we show that only 2.0% of the full training corpus of over 500,000 feature cases is needed to achieve correction results comparable to those using the entire training corpus, effectively reducing both the complexity of the training process and the learned correction model.

Why multiple document image binarizations improve OCR

Our previous work has shown that the error correction of optical character recognition (OCR) on degraded historical machine-printed documents is improved with the use of multiple information sources and multiple OCR hypotheses including from multiple document image binarizations. The contributions of this paper are in demonstrating how diversity among multiple binarizations makes those improvements to OCR accuracy possible. We demonstrate the degree and breadth to which the information required for correction is distributed across multiple binarizations of a given document image. Our analysis reveals that the sources of these corrections are not limited to any single binarization and that the full range of binarizations holds information needed to achieve the best result as measured by the word error rate (WER) of the final OCR decision. Even binarizations with high WERs contribute to improving the final OCR. For the corpus used in this research, fully 2.68% of all tokens are corrected using hypotheses not found in the OCR of the binarized image with the lowest WER. Further, we show that the higher the WER of the OCR overall, the more the corrections are distributed among all binarizations of the document image.

A synthetic document image dataset for developing and evaluating historical document processing methods

Document images accompanied by OCR output text and ground truth transcriptions are useful for developing and evaluating document recognition and processing methods, especially for historical document images. Additionally, research into improving the performance of such methods often requires further annotation of training and test data (e.g., topical document labels). However, transcribing and labeling historical documents is expensive. As a result, existing real-world document image datasets with such accompanying resources are rare and often relatively small. We introduce synthetic document image datasets of varying levels of noise that have been created from standard (English) text corpora using an existing document degradation model applied in a novel way. Included in the datasets is the OCR output from real OCR engines including the commercial ABBYY FineReader and the open-source Tesseract engines. These synthetic datasets are designed to exhibit some of the characteristics of an example real-world document image dataset, the Eisenhower Communiqu´es. The new datasets also benefit from additional metadata that exist due to the nature of their collection and prior labeling efforts. We demonstrate the usefulness of the synthetic datasets by training an existing multi-engine OCR correction method on the synthetic data and then applying the model to reduce word error rates on the historical document dataset. The synthetic datasets will be made available for use by other researchers.

Improving historical research by linking digital library information to a global genealogical database

Journals, letters, and other writings are of great value to historians and those who research their own family history; however, it can be difficult to find writings by specific people, and even harder to find what others wrote about them. We present a prototype web-based system that enables users to discover information about historical people (including their own ancestors) by linking digital library content to unique PersonIDs from a genealogical database. Users can contribute content such as scanned journals or information about where items can be found. They can also transcribe content and tag it with PersonIDs to identify who it is about. Additional features provide tools for users to explore historical contexts and relationships. These include the ability to tag places and to create a historical social network by specifying non-family relationships or by using a mechanism we call rosters to imply participation in some group or event.