Refine
Year of publication
Document Type
- Doctoral Thesis (58) (remove)
Language
- English (31)
- German (25)
- Multiple languages (2)
Keywords
- Agrarprodukt (1)
- Apfel (1)
- Architektur (3)
- Autonomous vessels (1)
- Autonomy (1)
- BIPV (1)
- Backstepping control (1)
- Bahnplanung (1)
- Beobachterentwurf (1)
- Bernstein Basis (1)
- Biomedical signals (1)
- Bodensee (1)
- COMSOL Multiphysics (1)
- Cauchon algorithm (1)
- Channel Coding (1)
- Code-based cryptography (1)
- Company Building (1)
- Compliance (1)
- Computer hardware (1)
- Computer security (1)
- Control Theory (1)
- Corporate Entrepreneurship (1)
- Corporate Responsibility (1)
- Corporate Social Responsibility (2)
- Corporate Venturing (1)
- Corporate venture (1)
- Corporate venture management (1)
- Crowdsourcing (1)
- Data encryption (Computer science) (1)
- Decision making (1)
- Digital and Analog Signal Processing, Automation (1)
- Drying (1)
- Dünne Schicht (1)
- EMV (1)
- Eletromobilität (1)
- Energieeffizientes Bauen (1)
- Entrepreneurship (1)
- Feuchtigkeitsmessung (1)
- Flash Memories (1)
- Gebäudedesign (1)
- Generalized Concatenated Code (1)
- Generalized concatenated codes (1)
- Hybrides Leistungsbündel (1)
- Höhenstrahlung (1)
- Incubation (1)
- Information Exchange (1)
- Innovation (1)
- Innovation Management (1)
- Integrity (1)
- Integrität in Unternehmen (1)
- Kaufentscheidung (1)
- Kollisionsschutz (1)
- Konvektionstrocknung (1)
- Körner, Edmund (1)
- Lebensmittel (1)
- Lebensmitteleinzelhandel (1)
- Lebensmittelqualität (2)
- Machine Learning (1)
- Machine learning (1)
- Magnetaktoren (1)
- Maschinelles Lernen (1)
- McEliece cryptosystem (1)
- Medical systems (1)
- Messtechnik (1)
- Microprocessors (1)
- Model predictive control (1)
- Modellierung (1)
- Multinationales Unternehmen (1)
- NLP (1)
- Nachhaltigkeit (1)
- Numerical Optimization (1)
- Oberflächenprüfung (1)
- Operations research (1)
- Optimization (1)
- Organisational Transformation (1)
- P2P (1)
- Post-quantum public-key cryptography (1)
- Preisbündelung (1)
- Principles of Blockchain Technology (1)
- Probabilistic Data Association (1)
- Product Development (1)
- Produkteigenschaft (1)
- Prozessmanagement (1)
- Prozessoptimierung (1)
- Quality retention (1)
- Random Matrices (1)
- Reichlin, Bruno (1)
- Requirements Engineering (1)
- Road Infrastructure Projects (1)
- SME (1)
- Sensorlose Folgeregelung (1)
- Sign regular matrix (1)
- Sign-regular matrix (1)
- Signal acquisition (1)
- Signal processing (1)
- Social entrepreneurship (1)
- Solaraktive Gebäudehülle (1)
- Solararchitektur (1)
- Startup Collaboration (1)
- Startup Investing (1)
- Strategic renewal (1)
- Strictly sign-regular matrix (1)
- Systemgeschäft (1)
- Target Tracking (1)
- Totally nonnegative matrix (1)
- Totally nonpositive matrix (1)
- Totally positive matrix (1)
- Tourismus (1)
- Trajectory tracking (1)
- Trocknung (3)
- Unbemanntes Fahrzeug (1)
- Unternehmensethik (1)
- Verbraucherverhalten (1)
- Verbundangebot (1)
- Vitamin C (1)
- Warenkennzeichnung (1)
- Wasserfahrzeug (1)
- aktiver Leistungsfilter (1)
- network traffic (1)
- selektive Signalanalyse (1)
- string matching (1)
- traffic classification (1)
- traffic obfuscation (1)
Institute
- Fakultät Bauingenieurwesen (1)
- Fakultät Elektrotechnik und Informationstechnik (1)
- Fakultät Informatik (3)
- Fakultät Maschinenbau (2)
- Fakultät Wirtschafts-, Kultur- und Rechtswissenschaften (3)
- Institut für Optische Systeme - IOS (3)
- Institut für Strategische Innovation und Technologiemanagement - IST (3)
- Institut für Systemdynamik - ISD (8)
- Institut für angewandte Thermo- und Fluiddynamik - IATF (1)
- Konstanz Institut für Corporate Governance - KICG (1)
Particularly for manufactured products subject to aesthetic evaluation, the industrial manufacturing process must be monitored, and visual defects detected. For this purpose, more and more computer vision-integrated inspection systems are being used. In optical inspection based on cameras or range scanners, only a few examples are typically known before novel examples are inspected. Consequently, no large data set of non-defective and defective examples could be used to train a classifier, and methods that work with limited or weak supervision must be applied. For such scenarios, I propose new data-efficient machine learning approaches based on one-class learning that reduce the need for supervision in industrial computer vision tasks. The developed novelty detection model automatically extracts features from the input images and is trained only on available non-defective reference data. On top of the feature extractor, a one-class classifier based on recent developments in deep learning is placed. I evaluate the novelty detector in an industrial inspection scenario and state-of-the-art benchmarks from the machine learning community. In the second part of this work, the model gets improved by using a small number of novel defective examples, and hence, another source of supervision gets incorporated. The targeted real-world inspection unit is based on a camera array and a flashing light illumination, allowing inline capturing of multichannel images at a high rate. Optionally, the integration of range data, such as laser or Lidar signals, is possible by using the developed targetless data fusion method.