Refine
Year of publication
Document Type
- Conference Proceeding (43)
- Article (24)
- Book (1)
Keywords
- automated optical inspection (1)
- line scan camera (1)
- particle filter (1)
- roller bearing (1)
- step detection (1)
- turn detection (1)
In this work we present a method to estimate an indoor position with the help of smartphone sensors and without any knowledge of absolute positioning systems like Wi-Fi signals. Our
system uses particle filtering to solve the recursive state estimation problem of finding the position of a pedestrian. We show how to integrate the information of the previous state into the weight update step and how the observation data can help within the state transition model. High positional accuracy can be achieved by only knowing that the pedestrian makes a foot step or changes her direction together with floor map information.
The Smartphone-Based Offline Indoor Location Competition at IPIN 2016: Analysis and Future Work
(2017)
Comparison of 2.4 GHz WiFi FTM- and RSSI-Based Indoor Positioning Methods in Realistic Scenarios
(2020)
3D Interaction Design: Increasing the stimulus-response correspondence by using stereoscopic vision
(2015)
This paper presents a hand-based interaction approach for grabbing and manipulating virtual objects within immersive virtual environments. Considering the stimulus-response of humans, interacting with objects is easier if depth information is provided by stereoscopic vision. However, most people are used to work in monoscopic virtual environments. Therefore, we propose a case study, to analyze if an interaction approach using stereoscopic vision, compared to a conventional monoscopic vision, results in a higher user acceptance and stimulus-response correspondence. Our interaction approach is based on a robust and fast running 3D hand- and head-tracking provided by Microsoft's Kinect. We present a simple non-collision-based method for grabbing and manipulating objects within a 3D virtual environment. By tracking the head, the user is able to move within the environment and interact depending on his current position. The use of stereoscopic vision allows a natural and instinctive hand interaction with the augmented object. Finally, the experimental results and studies show how the use of stereoscopic vision improves the mental mapping and thus the user acceptance and stimulus-response correspondence in consideration of using the here presented interaction approach.
An optical measuring method is presented, with which it is possible to measure and evaluate reflective cylindrical surfaces using a combination of deflectometry and subsequent reconstruction of the surface. The system is set up and tested on rolling elements of cylindrical roller bearings. However, it is not limited to this use case and can be applied to other cylindrical specular surfaces. The system distinguishes itself from existing test methods through the combination of high-resolution three-dimensional defect measurement with a very short recording time, and offers the possibility of introducing tolerance limits in the production of cylindrical specular surfaces. With this method, it is possible to record a defect with the dimensions of 1.3 mm by 1.8 mm within 5 s and to reconstruct the absolute depth. The resolution of the system is below 10 μm in both X- and Y-direction, and is therefore sufficiently accurate to detect typical surface defects such as scratches, dents, or deformations. To validate the measured values of the system, the results of an artificially generated 10.35 μm deep defect location were compared with those of a highly accurate mechanical stylus measurement.
Improving Statistical Object Recognition Approaches by a Parameterization of Normal Distributions
(2003)
Improving Statistical Object Recognition Approaches by a Parameterization of Normal Distributions
(2004)
As statistical approaches play an important role in object recognition, we present a novel approach which is based on object mod- els consisting of normal distributions for each training image. We show how to parameterize the mean vector and covariance matrix independently from the interpolation technique and formulate the classification and localization as a continuous optimization problem. This enables the computation of object poses which have never been seen during training. For interpolation, we present four different techniques which are compared in an experiment with real images. The results show the benefits of our method both in classification rate and pose estimation accuracy.
In the past decades most object recognition systems were based on passive approaches. But in the last few years a lot of research was done in the field of active object recognition. In this context there are several unique problems to be solved. One of them is how to fuse a series of images that might differ in their viewpoints. In this paper we present a well-founded approach for the fusion of multiple views based on a recursive density propagation method. It uses particle filters for solving the fusion in a continuous pose space. Furthermore we will show by means of a statistical object recognition system how to integrate such systems into our fusion approach. The experimental result will show, how the fusion can improve classification rates substantial, es-pecially for difficult conditions like heterogeneous background within real world environments.
In this paper we address the classification of 3–D objects that look similar from sev-eral sights and can only be distinguished from some certain viewpoints. For this purpose we combine a statistical appearance-based object recognition approach with an active viewpoint selection mechanism. For appearance-based object recognition lo-cal features are derived from wavelet multires-olution analysis. The recognition process is performed hierarchically in a statistical frame-work by a maximum likelihood estimation. Based on this result the active viewpoint se-lection mechanism chooses one further view that allows a reliable classification. Hereby the viewpoint selection mechanism can be trained unsupervised and represents the space of possible viewpoints continuously. Experimental results show that our approach is well suited for a reliable classifica-tion of similar looking objects only by one fur-ther view.
Active Sensing Strategies for Robotic Platforms, with an Application in Vision-Based Gripping
(2004)
For an effcient linguistic analysis of spoken queries a lot of domain specific knowledge is needed and usually has to be entered manually into the knowledge base of each domain. This makes the adaption of dialogue systems which base on explicit knowledge representation to new domains a very costly pro cedure. We use a frequency based statistical method combined with general hidden markov models in order to learn domain specific knowledge within a semantic network formalism. As a framework we use a dialogue system for German train timetable information. By means of
experiments we show that our statistical approach is not only able to reach, but even outperforms previous results with manually entered restrictions.