In this paper we propose a general, object-oriented software architecture for model-based visual tracking. The library is general purpose with respect to object model, estimated pose parameters, visual modalities employed, number of cameras and objects, and tracking methodology. The base class structure provides the necessary building blocks for implementing a wide variety of both known and nov...

In this paper, we present a tracking‐by‐multiple hypotheses framework to detect and track multiple vehicles accurately and precisely. The tracking‐by‐ multiple hypotheses framework consists of obstacle detection, vehicle recognition, visual tracking, global position tracking, data association and particle filtering. The multiple hypotheses are from obstacle detect...

Convolutional Neural Networks (CNNs) have demonstrated its great performance in various vision tasks, such as image classification [18] and object detection [8]. However, there are still some areas that are untouched, such as visual tracking. We believe that the biggest bottleneck of applying CNN for visual tracking is lack of training data. The power of CNN usually relies on huge (possible mil...

Surgical tool tracking is an important key functionality for many high-level tasks such as the visual guidance of surgical instruments or automated camera control. Readings from robot encoders and the kinematic chain are usually error prone in this kind of complex setup, but still allow for a coarse pose estimation of the instruments in image space. This information can be utilized to (re-)init...

In this paper, we propose and study a novel visual object tracking approach based on convolutional networks and recurrent networks. The proposed approach is distinct from the existing approaches to visual object tracking, such as filtering-based ones and tracking-by-detection ones, in the sense that the tracking system is explicitly trained off-line to track anonymous objects in a noisy environ...

In this paper, we present an overview of several visual tracking methods for industrial augmented reality applications. We show that no universal algorithm can deal with the large number of possible scenes, and that the different methods have to be seen as complementary approaches that all have their strengths and weaknesses. The main difficulty, then, consists in combining existing building bl...

Within the ITrackU project, a modular software architecture for model-based visual tracking and image understanding is being developed. The library is general-purpose with respect to object models, state-space parameters, visual modalities employed, number of cameras and targets, and tracking methodology. This provides the necessary building blocks for a seamless integration of a wide variety o...

This paper presents an attractive position-based visual sewoing approach for camera-in-hand robotic systems. The major contribution of this work is in devising an elegant and pragmatic approach for 3 0 visual tracking. The proposed Modijied Smith Predictor (MSP)-DeMenthon-Horaud (DH) visual sewoing system has shown to be reliable and yielded good target tracking performance. It d@er from the ot...

In this paper, we propose a visual tracking control method of a hand-eye robot for a moving target object with multiple feature points. The hand-eye robot is composed of a three degrees-offreedom planar manipulator and a single CCD camera that is mounted on the manipulator’s endeffector. The control objective is to keep all feature points of the target object around their desired coordinates on...

To achieve effective visual tracking, a robust feature representation composed of two separate components (i.e., feature learning and selection) for an object is one of the key issues. Typically, a common assumption used in visual tracking is that the raw video sequences are clear, while real-world data is with significant noise and irrelevant patterns. Consequently, the learned features may be...