Refine
Year of publication
Document Type
Keywords
- DTW (1)
- Parkinson (1)
- clustering (1)
- handwriting (1)
- on-line handwriting features (1)
- online strokes (1)
- parkinson disease (1)
- password identification (1)
- reduction feature (1)
- time series (1)
- writer verification (1)
Institute
Computergestützte Gesetze?
(1986)
Vergleich von Steuerungsmöglichkeiten anhand der Automatisierung eines Polykondensationsprozesses
(1983)
This work presents the conception and development of a software system to detect, track and count automatically moving objects in videos. The aim of the system is the automatically counting of moving turtles on a beach. To achieve a high recognition rate of the moving objects we consider three segmentation strategies (SubstractionGrayscale, SubstractionBinarization, Substraction Canny), two object identification methods (Grayscale Connected, FelzenwalbHuttenlocher) and two object recognition methods Nearest Object Distance, Certain Recognition Matching. In the preprocessing step of our system we select parameters like the interesting area and the object size. We use the segmentation process for the recognition of potential objects. After the identification of objects as turtles we track their movements. In the last step we count the turtles. In order to have a representative test number of moving objects we use videos with 1661 moving cars. The best results of 98.98 percent we reached with the combination Canny Edge Detection, Grayscale Connected, and Certain Region Matching Strategy.
Ziel des Beitrages ist es, das Potential von handschriftlichen Passworten für die Erhö- hung der Sicherheit beim Zugang zu Systemen zu untersuchen. Wir schlagen eine Mo- bilgerät-Server-Lösung mit verteilter Vorverarbeitung vor.
Untersucht werden die Erfolgsraten der korrekten Schreibererkennung bei der Verwen- dung nur eines Wortes, des Passwortes, der Reduktion der Anzahl der Passworte, die zum Trainieren des Klassifikationsalgorithmus verwendet werden, und der Reduktion der Parameter. Der implementierte Prototyp führt die Segmentierung auf dem Mobilgerät aus. Merkmalsextraktion und Klassifikation laufen als Server-Applikationen. Das Tool WEKA wird für die Ermittlung des geeignetsten Algorithmus für die Erkennung des Schreibers des Passwortes verwendet. Die Analysen werden mit 280 handschriftlichen Passworten durchgeführt. Davon sind 176 Originale und 104 Fälschungen. Als Mobilge- räte wurden ein HTC Desire mit Android 2.2 und ein Samsung Galaxy Ace mit Android 2.3.3 verwendet. Das vorgeschlagene System erkennt mit der Wahrscheinlichkeit von 96,87% den Schreiber des Passwortes, die False Acceptance Rate beträgt 12,5%, wenn acht Passworte jedes Schreibers für das Klassifikationsmodell verwendet werden.
We present a new method for detecting gait disorders according to their stadium using cluster methods for sensor data. 21 healthy and 18 Parkinson subjects performed the Time Up and Go test. The time series were segmented into separate steps. For the analysis the horizontal acceleration measured by a mobile sensor system was considered. We used Dynamic Time Warping and Hierarchical Custering to distinguish the stadiums. A specificity of 92% was achieved.
Im Rahmen der Studien „Defect Detection Study 21“ und „Dynamic Driving Study 21“ (DDS21) zur Reduzierung von Verkehrsunfällen im 21. Jahrhundert sollen die Möglich- keiten einer Videoauswertung von einer im Fahrzeug angebrachten Kamera untersucht werden. In einer ersten Version sollen nur Videodaten ohne Kenntnis anderer Fahr- zeugdaten dahingehend ausgewertet werden, ob sich fremde Fahrzeuge, deren Größe
und relative Geschwindigkeit zum Testfahrzeug ermitteln lassen
One of the biometric methods in authentication systems is the writer verification/identification using password handwriting. The main objective of this paper is to present a robust writer verification system by using cursive texts as well as block letter words. To evaluate the system, two datasets have been used. One of them is called Secure Password DB 150, which is composed of 150 users with 18 samples of single character words per user. Another dataset is public and called IAM online handwriting database, and it is composed of 220 users of cursive text samples. Each sample has been defined by a set of features, composed of 67 geometrical, statistical, and temporal features. In order to get more discriminative information, two feature reduction methods have been applied, Fisher Score and Info Gain Attribute Evaluation. Finally, the classification system has been implemented by hold-out cross validation and k-folds cross validation strategies for three different classifiers, K-NN, Naïve Bayes and Bayes Net classifiers. Besides, it has been applied for verification and identification approaches. The best results of 95.38% correct classification are achieved by using the k-nearest neighbor classifier for single character DB. A feature reduction by Info Gain Attribute Evaluation improves the results for Naïve Bayes Classifier to 98.34% for IAM online handwriting DB. It is concluded that the set of features and its reduction are a strong selection for the based-password handwritten writer identification in comparison with the state-of-the-art.