Bachelor Thesis

Nov 30, 20091 like12,056 views

This thesis describes a method to find a part of online data in an offline document. This method is able to find the offline document that belongs to the online data from a set of offline documents, or vice versa. In order to optimize the mapping between the online and the offline data, an optimal rotation and resizing of the online data is calculated. This is useful since it produces a better mapping between online and offline data, which makes several methods that are only applicable for online data available for offline data, and vice versa. Results show that this method can be used for finding the offline document that belongs to certain online data, since it succeeded in 98.07% of the cases for the used dataset. The results also show that computing the optimal rotation and resize factor significantly improves the mapping between online and offline data. This improvement is 6.56% for the used dataset.

Mapping online data on offline documents Artificial handwriting recognition Stefan Kennedie

Overview Introduction Method Experiment Discussion Conclusion

Introduction Online data Tablet Sequence of coordinates and pressure Offline data Digital scanner Bitmap with greyvalue or color

Properties of online and offline data Known Unknown Width of ink Unknown Known Location of pen known when not on paper Unknown Known Velocity and acceleration Unknown Known Order of writing Offline Online

Advantages of combining Applying online techniques to offline data and vice verca. e.g. segmentation Recognition improves

Combining – what’s the problem? Difference in resolution between tablet and digital scanner Allignment of paper on tablet and digital scanner Movement of paper (not investigated) Pen angle (unsolved)

Pen angle Difference in contact point Unsymmetric magnetic field

Goal Find correct match between offline and online data Create a match (mapping) between online and offline data that is as high as possible using: Rotation Resizing

Method – Overview Search for a query (created from online data) in the offline document In low (12.5%) resolution Mark locations with a good match Investigate these locations in detail, find best match using: Rotation Resizing

Method (1) – Query creation Create a query from online data Unique identification code Bresenham’s line algorithm 12.5% of original resolution of offline data

Method (1) – preprocessing offline data Upper half of first quadrant Remove noise Using Otsu Threshold algorithm Resize to 12.5% of original resolution

Method (1) – searching for query Find location with optimal match between query and offline document Compare query with offline document at all possible locations using: Sliding window Euclidean Distance Mapping for Matching Output: Match error Good match  small error Visualized as light pixel Bad match  big error Visualized as dark pixel

Method (1) – euclidean distance mapping example Offline document Query XOR image

Method (1) – processing match errors Remove locations with high error Remove error values Add surrounding locations 1 2 3

Method (2) – investigation in detail Find location, rotation and resize factor with best match.

Experiments Retrieval experiment – mapping in document level Optimal rotation and resizing experiment – mapping on pixel level Both done for three writers

Experiment 1 Find correct match between the online and offline data Uses only part one of the method, no rotation and resizing Compare all online data with all offline data (of one writer) Pairs with smallest match error belong to each other? Visual exploring the results

Experiment 1 – results In total 207 data pairs Algorithm found 203 correct matches (98.07%) % correct pixels for ‘match’ group: 81.08% % correct pixels for ‘no match’ group: 38.15% This difference of 42.93% is significant (p=.000)

Experiment 2 Compute the maximum % of correct pixels for all rotatoins and resizing factors for all data pairs.

Experiment 2 – results % of correct pixels without rotating and resizing: 80.72% % of correct pixels with optimal rotation and resizing factor: 87.28% Difference of 6.56% is significant (p=.000)

Discussion The method is very slow Alternative approach like Fast Foourier transformation The significance of the result depends on the data Large black regions in offline data Edge detection

Conclusion The method is useful for finding online-offline data pairs Succesfull in 98.07% The method is able to provide a better mapping between online and offline data 6.56% improvement Better results can be achieved using the pen angle

The document discusses Ghana's economic relationship with the European Union and the Economic Partnership Agreements (EPAs) between Ghana and the EU. It provides background on past trade agreements between Ghana and the EU. It outlines concerns that the EPAs could negatively impact Ghana's economy by eliminating tariffs and exposing local industries to more competition. However, proponents argue the EPAs could increase regional trade, help integrate Ghana's economy globally, and increase funding to support development. The document recommends the EU provide more support and capacity building to help Ghana benefit from the trade agreements.

Präsentation zu Bachelor-ArbeitDavid Baumeister

��

Präsentation Kolloquium Masterarbeit - BeispielMichael Fröse

��

Verteidigung Masterarbeitwruge

��

$Bachelor\'s Thesis Presentation$ $Bachelor\'s Thesis Presentation$

Bachelor\'s Thesis PresentationNatalia_Ladygina

��

Präsentation der Masterarbeit: Mobil Healt& Fitness - Anforderungsanalyse für...Nadine Edelmann

��

Heat discharge through forced cooling of phase change materials in concrete p...Kristin Kuckelkorn

��

This MSc thesis presented by Kristin Sophia Kuckelkorn examines using phase change materials in concrete to discharge heat through forced cooling, which allows for greater heat storage and reduces peak heating and cooling needs. The methodology section describes the experimental approach, while the results sections analyze how incorporating micronal PCM panels into concrete affects its properties and heat transfer rates when subjected to varying water flow rates and temperatures during forced cooling. The conclusion is that cast-in forced cooling can control heat drainage from PCM-enhanced concrete, with efficiency depending on climate conditions.

From Balanced Scorecard to Project Portfolio ManagementRoberto Toledo

��

The document discusses project portfolio management and alignment with organizational strategy. It introduces the Balanced Scorecard as a tool to help with strategic planning and implementation. The Balanced Scorecard provides a framework to classify objectives across four perspectives: financial, customer, internal processes, and learning and growth. It can then be used to identify projects and initiatives that will help achieve strategic goals.

짶�역축젵Ӛ�영메뉴얼Gori Communication

��

Business Process Maturity and Centers of ExcellenceSandy Kemsley

��

The document discusses BPM maturity models, the role of a BPM Center of Excellence (COE) in improving maturity, and how to build an effective COE. It describes the Object Management Group's maturity model and how a COE can help organizations progress from initial to innovative process maturity levels. A COE brings together best practices, training, knowledge sharing, and process governance to increase synergies between BPM projects. The document provides guidance on establishing the framework, methodology, team, and repositories needed for a successful COE.

Measuring Process Maturity: The Business Process Maturity ModelNathaniel Palmer

��

The document summarizes the Business Process Maturity Model (BPMM) which is used to assess an organization's process maturity. It discusses the five levels of the BPMM framework - initial/ad hoc, managed, standardized, predictable, and innovating. Higher maturity levels are associated with improved organizational performance, reduced rework, increased productivity and benefits. The BPMM provides a structured approach for organizations to measure, improve and manage their business processes.

UX 아카데미 오픈프로젝트 [ 배달의민족 - UX/UI 개선]RightBrain inc.

��

라이트브레인 UX 아카데미 22기 Open Project 로 이번에 소개해 드릴 결과물은 대표적인 배달앱인��‘배달의민족 – UX/UI 개선 프로젝트’입니다. 지난 UX 아카데미 22기 교육 기간 동안 배달의민족조가 만들어 낸 신선한 아이디어를 Rightbrain lab을 통해 공개해 드립니다.라이트브레인 아카데미 수강생들이 어떤 과정을 통해 아이디에이션 하고 또 인사이트와 컨셉을 도출했는지 궁금하시다면 지금 바로 슬라이드쉐어로 내용을 확인하세요. 3대 배달앱 시장 경쟁 과열, 새로운 전략의 필요성 배달 앱 업계에서는 매년 다양한 신종 배달앱들이 등장함에 따라 경쟁은 더욱 치열해지고 있습니다. 이러한 경쟁 과열 속에서 배달 앱 대표 타이틀을 굳건하게 유지하기 위한 배달의민족만의 차별화된 UX/UI 개선방안을 고민해 보았습니다.

Thesis Defense Presentationosideloc

��

Agile Product Manager/Product Owner Dilemma (PMEC)Enthiosys Inc

��

Li Yingying-- Master Thesis DefenseYingying Li

��

This thesis examines how LinkedIn China's sub-brand Chitu cultivates relationships with stakeholders in accordance with Chinese relationship and stakeholder management theories. It analyzes Chitu's relationship building strategies with key stakeholders like employees, users, and industry peers. Chitu focuses on triggering user self-expression and providing professional groups activities. Relationship cultivation is influenced by Chinese cultural concepts like guanxi, which Chitu leverages through reciprocity and trust. The thesis contributes to understanding how public relations theories apply in a Chinese context and offers implications for foreign companies to nurture stakeholder relationships in China.

107supplier relationshipbtecexpert

��

IBM would likely develop long-term, collaborative relationships with Keytronic focused on mutual benefits rather than individual transactions. Relationship purchasing requires focus on supplier retention through close cooperation and information sharing to improve factors like price, quality, and delivery. These types of relationships are useful as they allow IBM and Keytronic to work together innovatively as partners to achieve their goals. For the 'DimS' product, Paul should use a strategic alliance relationship strategy. This high-volume manufacturing industry benefits from a close, cooperative relationship where risks and rewards are shared. A strategic alliance allows for network sourcing through collaborative planning and is well-suited for the product's needs.

Najmul Hoda PhD Thesis Defence PresentationNajmul Hoda

��

Benefits management and transformational change, 10 January 2017 - SouthamptonAssociation for Project Management

��

Benefits Management: the essential ingredient for change, 10 Jan 2017. Southampton Transformational change is here to stay. We are living in an era of continuous transformation where standing still is simply not an option in today’s ultra-competitive and constantly changing business environments. This presentation describes the key drivers and management imperatives for successful transformational change in organisations. It shows how placing Benefits Management at the heart of change management directly addresses and enables success. In the main there are two key attributes of the Benefits Management methodology that help deliver successful business change. The first is a flexible framework which can be easily embedded across the entire organisational change structure. This provides powerful change management capabilities that focus on delivering the desired end results and outcomes for the business. Secondly and arguably most importantly, is the need to focus on the decision makers and data owners within the organisation. They are responsible for driving the change and associated benefits forward. This applies equally to; the senior responsible owner, the business change manager and benefit owners. Benefits Management done well will naturally draw people into the change process and can achieve game-changing results. It does this by fostering in people; joined-up thinking, good communications, readiness for change and a culture of shared goals and objectives. Today’s market drivers for constant business change don’t offer the luxury of choice. If there is new technology or a threat to our current or aspired market objectives, then organisations must rise to the change challenge or accept the inevitable consequences such as reduced market share or business failure. If directed to cut costs or do more for less, it should be done intelligently. Benefits management is essentially about making sure that the organisation has an unrelenting focus on delivering business value and not just traditional time, cost and quality outputs. In the real world, it’s about people and winning the hearts and minds of the people, that will deliver true success. On the theme of benefits management, the presentation impressed the need for measurable improvement but not at all costs. One of the challenges of benefits management is to identify which benefits are the ones to address, and not just because they happen to be measurable! Several useful frameworks and tools are recommended and referenced in the presentation. The presenters were Neil White, Chair of APM Benefits Management SIG and Merv Wyeth, Secretary. See also: http://bit.ly/2iykbXX

My PhD thesis presentation slidesMattia Bosio

��

Disparity Estimation by a Real Time Approximation AlgorithmCSCJournals

��

This document summarizes an approximation algorithm for real-time disparity estimation of stereo images. The algorithm shrinks the left and right images 3 times to reduce computational time and search area. Disparity is estimated from the shrunk images and then extrapolated to reconstruct the original disparity image. Experimental results on standard stereo images show the algorithm reduces computational time by 76.34% compared to traditional window-based methods, with acceptable accuracy. Some accuracy is lost due to pixel quantization during shrinking and extrapolation, but the fast estimation of dense disparity makes the algorithm useful for applications requiring real-time performance.

(Spring 2013) Signature Authentication Consistency Across DevicesInternational Center for Biometric Research

��

The purpose of this study is to determine the consistency of signatures and the reliability of verification across various devices. These devices include an iPad (with finger and stylus) and digitizers (back-lit and non-back-lit) for the purpose of electronic consent. This study was conducted using two to four (varied for each device and person) signatures per team member and then cross-comparing every combination of signature to determine the stability of using the algorithm across various devices with a given threshold.

More Related Content

What's hot (20)

Verteidigung der MasterThesischris2newz

��

Bachelor Thesis Presentationanderspurup

��

Präsentation zu Bachelor-ArbeitDavid Baumeister

��

Präsentation Kolloquium Masterarbeit - BeispielMichael Fröse

��

Verteidigung Masterarbeitwruge

��

$Bachelor\'s Thesis Presentation$ $Bachelor\'s Thesis Presentation$

Bachelor\'s Thesis PresentationNatalia_Ladygina

��

Präsentation der Masterarbeit: Mobil Healt& Fitness - Anforderungsanalyse für...Nadine Edelmann

��

Heat discharge through forced cooling of phase change materials in concrete p...Kristin Kuckelkorn

��

From Balanced Scorecard to Project Portfolio ManagementRoberto Toledo

��

짶�역축젵Ӛ�영메뉴얼Gori Communication

��

Business Process Maturity and Centers of ExcellenceSandy Kemsley

��

Measuring Process Maturity: The Business Process Maturity ModelNathaniel Palmer

��

UX 아카데미 오픈프로젝트 [ 배달의민족 - UX/UI 개선]RightBrain inc.

��

Thesis Defense Presentationosideloc

��

Agile Product Manager/Product Owner Dilemma (PMEC)Enthiosys Inc

��

Li Yingying-- Master Thesis DefenseYingying Li

��

107supplier relationshipbtecexpert

��

Najmul Hoda PhD Thesis Defence PresentationNajmul Hoda

��

Benefits management and transformational change, 10 January 2017 - SouthamptonAssociation for Project Management

��

My PhD thesis presentation slidesMattia Bosio

��

Verteidigung der MasterThesischris2newz

��

Bachelor Thesis Presentationanderspurup

��

Präsentation zu Bachelor-ArbeitDavid Baumeister

��

Präsentation Kolloquium Masterarbeit - BeispielMichael Fröse

��

Verteidigung Masterarbeitwruge

��

$Bachelor\'s Thesis Presentation$ $Bachelor\'s Thesis Presentation$

Bachelor\'s Thesis PresentationNatalia_Ladygina

��

Präsentation der Masterarbeit: Mobil Healt& Fitness - Anforderungsanalyse für...Nadine Edelmann

��

Heat discharge through forced cooling of phase change materials in concrete p...Kristin Kuckelkorn

��

From Balanced Scorecard to Project Portfolio ManagementRoberto Toledo

��

짶�역축젵Ӛ�영메뉴얼Gori Communication

��

Business Process Maturity and Centers of ExcellenceSandy Kemsley

��

Measuring Process Maturity: The Business Process Maturity ModelNathaniel Palmer

��

UX 아카데미 오픈프로젝트 [ 배달의민족 - UX/UI 개선]RightBrain inc.

��

Thesis Defense Presentationosideloc

��

Agile Product Manager/Product Owner Dilemma (PMEC)Enthiosys Inc

��

Li Yingying-- Master Thesis DefenseYingying Li

��

107supplier relationshipbtecexpert

��

Najmul Hoda PhD Thesis Defence PresentationNajmul Hoda

��

Benefits management and transformational change, 10 January 2017 - SouthamptonAssociation for Project Management

��

My PhD thesis presentation slidesMattia Bosio

��

Similar to Bachelor Thesis (20)

Disparity Estimation by a Real Time Approximation AlgorithmCSCJournals

��

(Spring 2013) Signature Authentication Consistency Across DevicesInternational Center for Biometric Research

��

BPI@BPM2011Jan Claes

��

This document discusses using an artificial immune system technique to merge computer log files from different sources for the purpose of process mining by identifying links between process instances across logs based on indicator factors like shared identifiers, attribute matching, timing differences, and allowing user input to refine the results. The proposed approach calculates scores for potential links based on these factors, combines them to identify the best match, and was tested on simulated and real data with mostly correct results, especially when unique identifiers were available.

Disparity Estimation Using A Color Segmentation V3thomaswangxin

��

The document discusses a segmentation-based stereo matching algorithm for disparity estimation. It first identifies challenges like textureless regions, depth discontinuities, and occlusions. It then outlines the algorithm which uses color segmentation, local matching, and polynomial modeling of disparity within segments. Disparity is refined through region merging. Experimental results on standard datasets show improvements over initial disparity estimation.

Paper 58 disparity-of_stereo_images_by_self_adaptive_algorithmMDABDULMANNANMONDAL

��

This document summarizes a research paper that proposes a new stereo matching algorithm called Self Adaptive Algorithm (SAA) to efficiently compute stereo correspondence or disparity maps from stereo images. SAA aims to improve matching speed by reducing the search zone and avoiding false matches through an adaptive search approach. It dynamically selects the search range based on previous matching results, reducing the range by 50% with each iteration. Experimental results on standard stereo datasets show that SAA outperforms other methods in terms of speed while maintaining accuracy, with processing speeds of 535 fps and 377 fps for different image pairs. SAA reduces computational time by 70.53-99.93% compared to other state-of-the-art methods.

Green Touch - ITS 12Consuelo Valdes

��

This document describes a study that evaluated the use of interactive surface technologies and mobile devices to help college students learn scientific concepts through collaborative inquiry. The study found that the technologies supported effective collaboration through opportunities for co-located data collection, discussion, and spatial interaction with data. Students were able to collect accurate and consistent field data using mobile devices. When using the surface technology, students collaborated well through turn-taking, role switching, and continuous coordination. The study provided implications for designing technologies that reduce complexity, support reflection, and allow for large data sets through spatial interaction across multiple devices.

Publication - The feasibility of gaze tracking for “mind reading” during searchA. LE

��

We perform thousands of visual searches every day, for example, when selecting items in a grocery store or when looking for a specific icon in a computer display. During search, our attention and gaze are guided toward visual features similar to those in the search target. This guidance makes it possible to infer information about the target from a searcher’s eye movements. The availability of compelling inferential algorithms could initiate a new generation of smart, gaze-controlled interfaces that deduce from their users’ eye movements the visual information for which they are looking. Here we address two fundamental questions: What are the most powerful algorithmic principles for this task, and how does their performance depend on the amount of available eye-movement data and the complexity of the target objects? While we choose a random-dot search paradigm for these analyses to eliminate contextual influences on search, the proposed techniques can be applied to the local feature vectors of any type of display. We present an algorithm that correctly infers the target pattern up to 66 times as often as a previously employed method and promises sufficient power and robustness for interface control. Moreover, the current data suggest a principal limitation of target inference that is crucial for interface design: If the target patterns exceed a certain spatial complexity level, only a subpattern tends to guide the observers' eye movements, which drastically impairs target inference.

My experimentBoshra Albayaty

��

This document outlines the PhD journey of Boshra F. Zopon Al_Bayaty in India. It discusses her coursework, research on knowledge discovery from web search, conferences attended both nationally and internationally, and her research contributions. Her research focused on using a "Master-Slave" model combining multiple supervised algorithms to improve word sense disambiguation and accuracy of over 70%. The document provides details on her research methodology, experiments conducted, results and opportunities for future work improving the model for other languages and applications. It demonstrates the various stages of her PhD journey and research progress.

Improvement of Spatial Data Quality Using the Data ConﬂationBeniamino Murgante

��

Iccsa stankuteha180611Beniamino Murgante

��

Data conflation is a process that optimizes spatial data quality by automatically merging heterogeneous geospatial datasets. It detects corresponding real-world features represented differently across datasets and combines their geometric and semantic information. For example, conflating datasets can identify a traffic roundabout modeled inconsistently and insert it accurately. By integrating multiple sources, conflation improves data completeness, positional accuracy, and temporal accuracy to produce an optimized dataset for applications.

Mutual informationMichael Munroe

��

1) The document describes the mutual information metric, which is used to evaluate the similarity between two images after an affine transformation. 2) It discusses evaluating the metric based on factors like execution time, speed, and quality of the optimization. It also examines the relationship between image size and computation time. 3) The document explores sub-pixel resolution with the metric and analyzing histograms of the original and transformed images to evaluate the mapping between them. It considers future directions like using the histogram to determine the next transform.

The human face of AI: how collective and augmented intelligence can help sol...Elena Simperl

��

This document summarizes a talk on how collective and augmented intelligence can help solve societal problems. It discusses how AI depends on human input, how collective intelligence benefits AI, and provides examples of using human computation and crowdsourcing to support disaster relief and conduct urban auditing. It also describes challenges in making crowdsourcing sustainable and assessing data quality, and emphasizes the need for iterative design of human-AI systems to bring together human, collective, and computational intelligence.

Fast Computational Four-Neighborhood Search Algorithm For Block matching Moti...IJERA Editor

��

The Motion estimation is an effective method for removing temporal redundancyfound in video sequence compression. Block Matching algorithm has been widely used in motion estimation and a number of fast algorithms have proposed to reduce the computational complexity of BMA. In this paper we propose a new search strategy for fast block matching based on Four-Neighborhood Search (FNS) and fast computational strategy, this new algorithm can significantly speed up the computation of the block matching by reducing the number of checked points and the time computational. Results have been shown that 89% to 93% of operations can be saved while maintaining the quality of video relative to full search algorithm.

Wang midterm-defenceZhipeng Wang

��

The document proposes four methods for improving object detection performance by combining different types of information. The first method uses common fate Hough transform to combine motion and appearance information. The second detects emergency indicators by fusing motion and appearance features. The third utilizes mutual information between image features using pyramid match score. The fourth method aims to detect objects with in-plane rotations by analyzing votes from different keypoints. The methods are evaluated on various datasets and aim to better utilize additional information for more accurate detection.

Thesis Presentationnirvdrum

��

The document discusses a study evaluating different user feedback systems for search engines. It describes previous work using implicit and explicit user feedback to improve search result relevance. The study collected both voluntary and mandatory feedback from users in controlled and uncontrolled search scenarios. Decision trees were built from the collected data and showed that voluntary feedback systems achieved higher classification accuracy than mandatory systems, supporting the hypothesis that voluntary feedback is higher quality.

FinalReportAnıl Ulaş KOÇAK

��

This document compares four recent boundary detection methods: Boundary Detection with Sketch Tokens, Crisp Boundary Detection Using Pointwise Mutual Information, Oriented Edge Forests for Boundary Detection, and Fast Edge Detection Using Structured Forests. It evaluates these methods on the Berkeley Segmentation Dataset using precision-recall metrics to determine which method achieves the best balance of high precision and high recall. It also considers the time complexity and simplicity of each method to help researchers select the most appropriate approach for their needs.

PEMF2_SDM_2012_AliMDO_Lab

��

Approximation models (or surrogate models) provide an efficient substitute to expen- sive physical simulations and an efficient solution to the lack of physical models of system behavior. However, it is challenging to quantify the accuracy and reliability of such approximation models in a region of interest or the overall domain without additional system evaluations. Standard error measures, such as the mean squared error, the cross-validation error, and the Akaikes information criterion, provide limited (often inadequate) information regarding the accuracy of the final surrogate. This paper introduces a novel and model independent concept to quantify the level of errors in the function value estimated by the final surrogate in any given region of the design domain. This method is called the Re- gional Error Estimation of Surrogate (REES). Assuming the full set of available sample points to be fixed, intermediate surrogates are iteratively constructed over a sample set comprising all samples outside the region of interest and heuristic subsets of samples inside the region of interest (i.e., intermediate training points). The intermediate surrogate is tested over the remaining sample points inside the region of interest (i.e., intermediate test points). The fraction of sample points inside region of interest, which are used as intermediate training points, is fixed at each iteration, with the total number of iterations being pre-specified. The estimated median and maximum relative errors within the region of interest for the heuristic subsets at each iteration are used to fit a distribution of the median and maximum error, respectively. The estimated statistical mode of the median and the maximum error, and the absolute maximum error are then represented as functions of the density of intermediate training points, using regression models. The regression models are then used to predict the expected median and maximum regional errors when all the sample points are used as training points. Standard test functions and a wind farm power generation problem are used to illustrate the effectiveness and the utility of such a regional error quantification method.

One dimensional vector based patternijcsit

��

Template matching is a basic method in image analysis to extract useful information from images. In this paper, we suggest a new method for pattern matching. Our method transform the template image from two dimensional image into one dimensional vector. Also all sub-windows (same size of template) in the reference image will transform into one dimensional vectors. The three similarity measures SAD, SSD, and Euclidean are used to compute the likeness between template and all sub-windows in the reference image to find the best match. The experimental results show the superior performance of the proposed method over the conventional methods on various template of different sizes.

Surface Data Capture Image MatchingRASHID JAVED

��

This document summarizes a study on automatic image matching using area-based correlation. The study developed a procedure to automatically determine conjugate points in epipolar image pairs using template matching. Different template sizes and correlation thresholds were tested. While the method worked well in textured areas, it struggled in homogeneous and shadowed regions with constant gray values. Increasing the template size improved matching but also introduced more geometric distortion. The results provided correlated points and a parallax map but were unable to generate a surface model.

Tracking Robustness and Green View Index Estimation of Augmented and Diminish...Tomohiro Fukuda

��

Disparity Estimation by a Real Time Approximation AlgorithmCSCJournals

��

(Spring 2013) Signature Authentication Consistency Across DevicesInternational Center for Biometric Research

��

BPI@BPM2011Jan Claes

��

Disparity Estimation Using A Color Segmentation V3thomaswangxin

��

Paper 58 disparity-of_stereo_images_by_self_adaptive_algorithmMDABDULMANNANMONDAL

��

Green Touch - ITS 12Consuelo Valdes

��

Publication - The feasibility of gaze tracking for “mind reading” during searchA. LE

��

My experimentBoshra Albayaty

��

Improvement of Spatial Data Quality Using the Data ConﬂationBeniamino Murgante

��

Iccsa stankuteha180611Beniamino Murgante

��

Mutual informationMichael Munroe

��

The human face of AI: how collective and augmented intelligence can help sol...Elena Simperl

��

Fast Computational Four-Neighborhood Search Algorithm For Block matching Moti...IJERA Editor

��

Wang midterm-defenceZhipeng Wang

��

Thesis Presentationnirvdrum

��

FinalReportAnıl Ulaş KOÇAK

��

PEMF2_SDM_2012_AliMDO_Lab

��

One dimensional vector based patternijcsit

��

Surface Data Capture Image MatchingRASHID JAVED

��

Tracking Robustness and Green View Index Estimation of Augmented and Diminish...Tomohiro Fukuda

��

Recently uploaded (20)

The Constitution, Government and Law making bodies .saanidhyapatel09

��

Year 10 The Senior Phase Session 3 Term 1.pptxmansk2

��

Fuel part 1.pptx........................ksbhattadcm

��

Adventure Activities Final By H R Gohil SirGUJARATCOMMERCECOLLE

��

POWERPOINT-PRESENTATION_DM-NO.017-S.2025.pptxMarilenQuintoSimbula

��

EDL 290F Week 3 - Mountaintop Views (2025).pdfLiz Walsh-Trevino

��

The Dravidian Languages: Tamil, Telugu, Kannada, Malayalam, Brahui, Kuvi, TuluDrIArulAram

��

A PPT on the First Three chapters of Wings of FireBeena E S

��

Database population in Odoo 18 - Odoo slidesCeline George

��

Digital Tools with AI for e-Content Development.pptxDr. Sarita Anand

��

English 4 Quarter 4 Week 4 Classroom ObsNerissaMendez1

��

How to Modify Existing Web Pages in Odoo 18Celine George

��

TPR Data strategy 2025 (1).pdf Data strategyHenry Tapper

��

Mate, a short story by Kate Grenvile.pptxLiny Jenifer

��

Blind Spots in AI and Formulation Science Knowledge Pyramid (Updated Perspect...Ajaz Hussain

��

This presentation delves into the systemic blind spots within pharmaceutical science and regulatory systems, emphasizing the significance of "inactive ingredients" and their influence on therapeutic equivalence. These blind spots, indicative of normalized systemic failures, go beyond mere chance occurrences and are ingrained deeply enough to compromise decision-making processes and erode trust. Historical instances like the 1938 FD&C Act and the Generic Drug Scandals underscore how crisis-triggered reforms often fail to address the fundamental issues, perpetuating inefficiencies and hazards. The narrative advocates a shift from reactive crisis management to proactive, adaptable systems prioritizing continuous enhancement. Key hurdles involve challenging outdated assumptions regarding bioavailability, inadequately funded research ventures, and the impact of vague language in regulatory frameworks. The rise of large language models (LLMs) presents promising solutions, albeit with accompanying risks necessitating thorough validation and seamless integration. Tackling these blind spots demands a holistic approach, embracing adaptive learning and a steadfast commitment to self-improvement. By nurturing curiosity, refining regulatory terminology, and judiciously harnessing new technologies, the pharmaceutical sector can progress towards better public health service delivery and ensure the safety, efficacy, and real-world impact of drug products.

How to attach file using upload button Odoo 18Celine George

��

APM People Interest Network Conference - Tim Lyons - The neurological levels ...Association for Project Management

��

APM People Interest Network Conference 2025 -Autonomy, Teams and Tension: Projects under stress -Tim Lyons -The neurological levels of team-working: Harmony and tensions With a background in projects spanning more than 40 years, Tim Lyons specialised in the delivery of large, complex, multi-disciplinary programmes for clients including Crossrail, Network Rail, ExxonMobil, Siemens and in patent development. His first career was in broadcasting, where he designed and built commercial radio station studios in Manchester, Cardiff and Bristol, also working as a presenter and programme producer. Tim now writes and presents extensively on matters relating to the human and neurological aspects of projects, including communication, ethics and coaching. He holds a Master’s degree in NLP, is an NLP Master Practitioner and International Coach. He is the Deputy Lead for APM’s People Interest Network. Session | The Neurological Levels of Team-working: Harmony and Tensions Understanding how teams really work at conscious and unconscious levels is critical to a harmonious workplace. This session uncovers what those levels are, how to use them to detect and avoid tensions and how to smooth the management of change by checking you have considered all of them.

N.C. DPI's 2023 Language Diversity BriefingMebane Rash

��

Kaun TALHA quiz Finals -- El Dorado 2025Conquiztadors- the Quiz Society of Sri Venkateswara College

��

CRITICAL THINKING AND NURSING JUDGEMENT.pptxPoojaSen20

��

The Constitution, Government and Law making bodies .saanidhyapatel09

��

Year 10 The Senior Phase Session 3 Term 1.pptxmansk2

��

Fuel part 1.pptx........................ksbhattadcm

��

Adventure Activities Final By H R Gohil SirGUJARATCOMMERCECOLLE

��

POWERPOINT-PRESENTATION_DM-NO.017-S.2025.pptxMarilenQuintoSimbula

��

EDL 290F Week 3 - Mountaintop Views (2025).pdfLiz Walsh-Trevino

��

The Dravidian Languages: Tamil, Telugu, Kannada, Malayalam, Brahui, Kuvi, TuluDrIArulAram

��

A PPT on the First Three chapters of Wings of FireBeena E S

��

Database population in Odoo 18 - Odoo slidesCeline George

��

Digital Tools with AI for e-Content Development.pptxDr. Sarita Anand

��

English 4 Quarter 4 Week 4 Classroom ObsNerissaMendez1

��

How to Modify Existing Web Pages in Odoo 18Celine George

��

TPR Data strategy 2025 (1).pdf Data strategyHenry Tapper

��

Mate, a short story by Kate Grenvile.pptxLiny Jenifer

��

Blind Spots in AI and Formulation Science Knowledge Pyramid (Updated Perspect...Ajaz Hussain

��

How to attach file using upload button Odoo 18Celine George

��

APM People Interest Network Conference - Tim Lyons - The neurological levels ...Association for Project Management

��

N.C. DPI's 2023 Language Diversity BriefingMebane Rash

��

Kaun TALHA quiz Finals -- El Dorado 2025Conquiztadors- the Quiz Society of Sri Venkateswara College

��

CRITICAL THINKING AND NURSING JUDGEMENT.pptxPoojaSen20

��

Bachelor Thesis

1. Mapping online data on offline documents Artificial handwriting recognition Stefan Kennedie

2. Overview Introduction Method Experiment Discussion Conclusion

3. Introduction Online data Tablet Sequence of coordinates and pressure Offline data Digital scanner Bitmap with greyvalue or color

4. Properties of online and offline data Known Unknown Width of ink Unknown Known Location of pen known when not on paper Unknown Known Velocity and acceleration Unknown Known Order of writing Offline Online

5. Advantages of combining Applying online techniques to offline data and vice verca. e.g. segmentation Recognition improves

6. Combining – what’s the problem? Difference in resolution between tablet and digital scanner Allignment of paper on tablet and digital scanner Movement of paper (not investigated) Pen angle (unsolved)

7. Pen angle Difference in contact point Unsymmetric magnetic field

8. Goal Find correct match between offline and online data Create a match (mapping) between online and offline data that is as high as possible using: Rotation Resizing

9. Method – Overview Search for a query (created from online data) in the offline document In low (12.5%) resolution Mark locations with a good match Investigate these locations in detail, find best match using: Rotation Resizing

10. Method (1) – Query creation Create a query from online data Unique identification code Bresenham’s line algorithm 12.5% of original resolution of offline data

11. Method (1) – preprocessing offline data Upper half of first quadrant Remove noise Using Otsu Threshold algorithm Resize to 12.5% of original resolution

12. Method (1) – searching for query Find location with optimal match between query and offline document Compare query with offline document at all possible locations using: Sliding window Euclidean Distance Mapping for Matching Output: Match error Good match  small error Visualized as light pixel Bad match  big error Visualized as dark pixel

13. Method (1) – euclidean distance mapping example Offline document Query XOR image

14. Method (1) – processing match errors Remove locations with high error Remove error values Add surrounding locations 1 2 3

15. Method (2) – investigation in detail Find location, rotation and resize factor with best match.

16. Experiments Retrieval experiment – mapping in document level Optimal rotation and resizing experiment – mapping on pixel level Both done for three writers

17. Experiment 1 Find correct match between the online and offline data Uses only part one of the method, no rotation and resizing Compare all online data with all offline data (of one writer) Pairs with smallest match error belong to each other? Visual exploring the results

18. Experiment 1 – results In total 207 data pairs Algorithm found 203 correct matches (98.07%) % correct pixels for ‘match’ group: 81.08% % correct pixels for ‘no match’ group: 38.15% This difference of 42.93% is significant (p=.000)

19. Experiment 2 Compute the maximum % of correct pixels for all rotatoins and resizing factors for all data pairs.

20. Experiment 2 – results % of correct pixels without rotating and resizing: 80.72% % of correct pixels with optimal rotation and resizing factor: 87.28% Difference of 6.56% is significant (p=.000)

21. Discussion The method is very slow Alternative approach like Fast Foourier transformation The significance of the result depends on the data Large black regions in offline data Edge detection

22. Conclusion The method is useful for finding online-offline data pairs Succesfull in 98.07% The method is able to provide a better mapping between online and offline data 6.56% improvement Better results can be achieved using the pen angle

23. Questions ?

�ݺ�ߣ

Bachelor Thesis

Recommended

More Related Content

What's hot (20)

Similar to Bachelor Thesis (20)

Recently uploaded (20)

Bachelor Thesis