Search results

The purpose of this paper is to develop a text extraction tool for scanned documents that would extract text and build the keywords corpus and key phrases corpus for the document without manual intervention.

For text extraction from scanned documents, a Web-based optical character recognition (OCR) tool was developed. OCR is a well-established technology, so to develop the OCR, Microsoft Office document imaging tools were used. To account for the commonly encountered problem of skew being introduced, a method to detect and correct the skew introduced in the scanned documents was developed and integrated with the tool. The OCR tool was customized to build keywords and key phrases corpus for every document.

The developed tool was evaluated using a 100 document corpus to test the various properties of OCR. The tool had above 99 per cent word read accuracy for text only image documents. The customization of the OCR was tested with samples of Microfiches, sample of Journal pages from back volumes and samples from newspaper clips and the results are discussed in the summary. The tool was found to be useful for text extraction and processing.

The scanned documents are converted to keywords and key phrases corpus. The tool could be used to build metadata for scanned documents without manual intervention.

The tool is used to convert unstructured data (in the form of image documents) to structured data (the document is converted into keywords, and key phrases database). In addition, the image document is converted to editable and searchable document.

The purpose of this paper is to develop an online smoothing zero-velocity-update (ZUPT) method that helps achieve smooth estimation of human foot motion for the ZUPT-aided inertial pedestrian navigation system.

The smoothing ZUPT is based on a Rauch–Tung–Striebel (RTS) smoother, using a six-state Kalman filter (KF) as the forward filter. The KF acts as an indirect filter, which allows the sensor measurement error and position error to be excluded from the error state vector, so as to reduce the modeling error and computational cost. A threshold-based strategy is exploited to verify the detected ZUPT periods, with the threshold parameter determined by a clustering algorithm. A quantitative index is proposed to give a smoothness estimate of the position data.

Experimental results show that the proposed method can improve the smoothness, robustness, efficiency and accuracy of pedestrian navigation.

Because of the chosen smoothing algorithm, a delay no longer than one gait cycle is introduced. Therefore, the proposed method is suitable for applications with soft real-time constraints.

The paper includes implications for the smooth estimation of most types of pedal locomotion that are achieved by legged motion, by using a sole foot-mounted commercial-grade inertial sensor.

This paper helps realize smooth transitions between swing and stance phases, helps enable continuous correction of navigation errors during the whole gait cycle, helps achieve robust detection of gait phases and, more importantly, requires lower computational cost.

Automated skin lesion analysis plays a vital role in early detection. Having relatively small-sized imbalanced skin lesion datasets impedes learning and dominates research in automated skin lesion analysis. The unavailability of adequate data poses difficulty in developing classification methods due to the skewed class distribution.

Boosting-based transfer learning (TL) paradigms like Transfer AdaBoost algorithm can compensate for such a lack of samples by taking advantage of auxiliary data. However, in such methods, beneficial source instances representing the target have a fast and stochastic weight convergence, which results in “weight-drift” that negates transfer. In this paper, a framework is designed utilizing the “Rare-Transfer” (RT), a boosting-based TL algorithm, that prevents “weight-drift” and simultaneously addresses absolute-rarity in skin lesion datasets. RT prevents the weights of source samples from quick convergence. It addresses absolute-rarity using an instance transfer approach incorporating the best-fit set of auxiliary examples, which improves balanced error minimization. It compensates for class unbalance and scarcity of training samples in absolute-rarity simultaneously for inducing balanced error optimization.

Promising results are obtained utilizing the RT compared with state-of-the-art techniques on absolute-rare skin lesion datasets with an accuracy of 92.5%. Wilcoxon signed-rank test examines significant differences amid the proposed RT algorithm and conventional algorithms used in the experiment.

Experimentation is performed on absolute-rare four skin lesion datasets, and the effectiveness of RT is assessed based on accuracy, sensitivity, specificity and area under curve. The performance is compared with an existing ensemble and boosting-based TL methods.

This paper provides an introduction to the use of n‐grams in textual information systems, where an n‐gram is a string of n, usually adjacent, characters extracted from a section of continuous text. Applications that can be implemented efficiently and effectively using sets of n‐grams include spelling error detection and correction, query expansion, information retrieval with serial, inverted and signature files, dictionary look‐up, text compression, and language identification.

This study aims to design an optimized algorithm for low-cost pedestrian navigation system (PNS) to correct the heading drift and altitude error, thus achieving high-precise pedestrian location in both two-dimensional (2-D) and three-dimensional (3-D) space.

A novel heading correction algorithm based on smoothing filter at the terminal of zero velocity interval (ZVI) is proposed in the paper. This algorithm adopts the magnetic sensor to calculate all the heading angles in the ZVI and then applies a smoothing filter to obtain the optimal heading angle. Furthermore, heading correction is executed at the terminal moment of ZVI. Meanwhile, an altitude correction algorithm based on step height constraint is proposed to suppress the altitude channel divergence of strapdown inertial navigation system by using the step height as the measurement of the Kalman filter.

The verification experiments were carried out in 2-D and 3-D space to evaluate the performance of the proposed pedestrian navigation algorithm. The results show that the heading drift and altitude error were well corrected. Meanwhile, the path calculated by the novel algorithm has a higher match degree with the reference trajectory, and the positioning errors of the 2-D and 3-D trajectories are both less than 0.5 per cent.

Besides zero velocity update, another two problems, namely, heading drift and altitude error in the PNS, are solved, which ensures the high positioning precision of pedestrian in indoor and outdoor environments.

Simple and affordable systems are described to assist wheelchair users in steering their wheelchairs across sloping ground. The systems can be attached to many standard powered wheelchairs. Wheelchairs often steer by having two swivelling caster wheels but problems with this configuration occur when a wheelchair is driven along sloping ground because the casters can swivel in the direction of the slope. Gravity then causes the wheelchair to start an unwanted turn or ‘veer’ and the chair goes in an unintended direction. This situation is exacerbated for switch users, as switches cannot provide fine control to trim and compensate. Early experiments demonstrated that calibrating wheelchair controllers for straight‐line balance and optimising motor‐compensation did not solve this problem. Caster angle was selected to provide feedback to the wheelchair controllers. At the point when veer is first detected, a wheelchair has already begun to alter course and the job of the correction system is to minimise this drift from the desired course. A rolling road was created as an assessment tool and trials with both the test bed and in real situations were conducted to evaluate the new systems. The small swivel detector that was created could be successfully attached to caster swivel bearings. The new system was successful, robust and was not affected by changeable parameters. Although primarily intended for switch users, the methods can be applied to users with proportional controls.

An increasing volume of historical text is being converted into machine‐readable form so as to allow database searches to be carried out. The age of the material in these databases means that they contain many spellings that are different from those used today. This characteristic means that, once the databases become available for general online access, users will need to be familiar with all of the possible historical spellings for their topic of interest if a search is to be carried out successfully. This paper investigates the use of computational techniques that have been developed for the correction of spelling errors to identify historical spellings of a user's search terms. Two classes of spelling correction method are tested, these being the reverse error and phonetic coding methods. Experiments with words from the Hartlib Papers Collection show that these methods can correctly identify a large number of historical forms of modern‐day word spellings.

The purpose of this paper is to develop an electronic solution to effectively lock swivelling wheel steering positions to driver‐control. Simple and affordable systems are described to assist forklift users in steering their walkie type forklifts or pallet jacks across sloping ground.

A rolling road was created as an assessment tool and trials with both the test bed and in real situations were conducted to evaluate the new systems. The small swivel detector that was created could be successfully attached to swivelling wheel swivel bearings.

The new system was successful, robust and was not affected by changeable parameters. The simple systems assisted hand truck operators in steering their forklifts across sloping ground without veering off course. The systems overcame the problems associated with forklifts that steer using two swivelling wheels and meant that less work was required from hand truck operators as their forklifts tended to travel in the desired direction

Experiments demonstrated that calibrating forklift controllers for straight‐line balance and optimizing motor‐compensation did not solve this problem. Instead, swivelling wheel angle was selected to provide feedback. At the point when veer is first detected, a forklift has already begun to alter course and the job of the correction system is to minimize this drift from the desired course.

The forklifts and pallet jacks often steer by having swivelling wheels but problems with this configuration occur when a forklift is driven along sloping ground because they can swivel in the direction of the slope. Gravity then causes the forklift or pallet jack to start an unwanted turn or “veer” and the vehicle goes in an unintended direction. This situation is exacerbated for vehicles with switch controls, as switches cannot provide fine control to trim and compensate.

Each year in the United States, over 100 employees are killed and 36,000 are seriously injured in accidents involving forklift trucks and pallet carriers. This is the second leading cause of occupational fatalities in “industrial” type workplaces. The research aims to make the use of this type of equipment safer and the systems can be attached to many standard forklifts and pallet jacks.

Quality control charts are statistical process control tools aimed at monitoring a (manufacturing) process to detect any deviations from normal operation and to aid in process diagnosis and correction. The information presented on the chart is a key to the successful implementation of a quality process correction system. Pattern recognition methodology has been pursued to identify unnatural behaviour on quality control charts. This approach provides the ability to utilize patterning information of the chart and to track back the root causes of process deviation, thus facilitating process diagnosis and maintenance. Presents analysis and development of a statistical pattern recognition system for the explicit identification of unnatural patterns on control charts. Develops a set of statistical pattern recognizers based on the likelihood ratio approach and on correlation analysis. Designs and implements a training algorithm to maximize the probability of identifying unnatural patterns, and presents a classification procedure for real‐time operation. Demonstrates the system performance using a set of newly defined measures, and obtained results based on extensive experiments illustrate the power and usefulness of the statistical approach for automating unnatural pattern detection on control charts.

The purpose of this paper is to provide much needed data to staff working with archival digitization on cost and benefit of visual checks during quality control workflows, and to encourage those in the field of digitization to take a data-driven approach to planning and workflow development as they transition into large-scale digitization.

This is a case study of a cost benefit analysis at the Triangle Research Libraries Network. Data were tracked on time spent performing visual checks compared to scanning production and error type/discovery rates for the consortial grant “Content, context, and capacity: a collaborative large-scale digitization project on the long civil rights movement in North Carolina”.

Findings show that 85 percent of time was spent scanning and 15 percent was spent on quality control with visual checks of every scan. Only one error was discovered for every 223 scans reviewed (0.4 percent of scans). Of the six types of error identified, only half cause critical user experience issues. Of all errors detected, only 32 percent fell into the critical category. One critical error was found for every 700 scans (0.1 percent of scans). If all the time spent performing visual checks were instead spent on scanning, production would have increased by 18 percent. Folders with 100 or more scans comprised only 11.5 percent of all folders and 37 percent of folders in this group contained errors (for comparison, only 8 percent of folders with 50 or more scans contained errors). Additionally, 52 percent of all critical errors occurred in these folders. The errors in larger folders represented 30 percent of total errors, and performing visual checks on the large folders required 32 percent of all visual check time.

The data gathered during this research can be repurposed by others wishing to consider or conduct cost benefit analysis of visual check workflows for large-scale digitization.

To the authors' knowledge, this is the only available dataset on rate of error detection and error type compared to time spent on quality control visual checks in digitization.