PANORAMA 2014 Abstracts


Full Papers
Paper Nr: 1
Title:

Mutation Detection System for Actualizing Traffic Sign Inventories

Authors:

Lykele Hazelhoff, Ivo Creusen and Peter H. N. de With

Abstract: Road safety is influenced by the adequate placement of traffic signs. As the visibility of road signs degrades over time due to e.g. aging, vandalism or vegetation coverage, sign maintenance is required to preserve a high road safety. This is commonly performed based on inventories of traffic signs, which should be conducted periodically, as road situations may change and the visibility of signs degrades over time. These inventories are created efficiently from street-level images by (semi-)automatic road sign recognition systems, employing computer vision techniques for sign detection and classification. Instead of periodically repeating the complete surveying process, these automated sign recognition systems enable re-identification of the previously found signs. This results in the highlighting of changed situations, enabling specific manual validation of these cases. This paper presents a mutation detection approach for semi-automatic updating of traffic sign inventories, together with a case study to assess the practical usability of such an approach. Our system re-identifies 94.8% of the unchanged signs, thereby resulting in a significant reduction of the manual effort required for the semi-automated actualization of the inventory. As the amount of changes equals to 16:9% of the already existing signs, this study also clearly shows the economic relevance and usefulness of periodic updating road sign surveys.

Paper Nr: 2
Title:

A Full Reference Video Quality Measure based on Motion Differences and Saliency Maps Evaluation

Authors:

B. Ortiz-Jaramillo, A. Kumcu, L. Platisa and W. Philips

Abstract: While subjective assessment is recognized as the most reliable means of quantifying video quality, objective assessment has proven to be a desirable alternative. Existing video quality indices achieve reasonable prediction of human quality scores, and are able to well predict quality degradation due to spatial distortions but not so well those due to temporal distortions. In this paper, we propose a perception-based quality index in which the novelty is the direct use of motion information to extract temporal distortions and to model the human visual attention. Temporal distortions are computed from optical flow and common vector metrics. Results of psychovisual experiments are used for modeling the human visual attention. Results show that the proposed index is competitive with current quality indices presented in the state of art. Additionally, the proposed index is much faster than other indices also including a temporal distortion measure.

Paper Nr: 3
Title:

Online Non-rigid Structure-from-Motion based on a Keyframe Representation of History

Authors:

Simon Donné, Ljubomir Jovanov , Bart Goossens, Wilfried Philips and Aleksandra Pižurica

Abstract: Non-rigid structure-from-motion in an on-line setting holds many promises for useful applications, and off-line reconstruction techniques are already very advanced. Literature has only recently started focusing on on-line reconstruction, with only a handful of existing techniques available. Here we propose a novel method of history representation which utilizes the advances in off-line reconstruction. We represent the history as a set of keyframes, a representative subset of all past frames. This history representation is used as side-information in the estimation of individual frames. We expand the history as previously unseen frames arrive and compress it again when its size grows too large. We evaluate the proposed method on some test sequences, focusing on a human face in a conversation. While on-line algorithms can never perform as well as off-line methods as they have less information available, our method compares favourably to the state of the art off-line methods.

Paper Nr: 4
Title:

Retrieval System for Person Re-identification

Authors:

Sławomir Bąk, François Brémond, Vasanth Bathrinarayanan, Alessandro Capra, Davide Giacalone, Giuseppe Messina and Antonio Buemi

Abstract: This paper addresses the problem of person re-identification and its application to a real world scenario. We introduce a retrieval system that helps a human operator in browsing a video content. This system is designed for determining whether a given person of interest has already appeared over a network of cameras. In contrast to most of state of the art approaches we do not focus on searching the best strategy for feature matching between camera pairs, but we explore techniques that can perform relatively well in a whole network of cameras. This work is devoted to analyze current state of the art algorithms and technologies, which are currently available on the market. We examine whether current techniques may help a human operator in solving this challenging task. We evaluate our system on the publicly available dataset and demonstrate practical advantages of the proposed solutions.

Paper Nr: 6
Title:

On Multi-view Texture Mapping of Indoor Environments using Kinect Depth Sensors

Authors:

Luat Do, Lingni Ma, Egor Bondarev and Peter H. N. de With

Abstract: At present, research on reconstruction and coloring of 3D models is growing rapidly due to increasing availability of low-cost 3D sensing systems. In this paper, we explore coloring of triangular mesh models with multiple color images by employing a multi-view texture mapping approach. The fusion of depth and color vision data is complicated by 3D modeling and multi-viewpoint registration inaccuracies. In addition, the large amount of camera viewpoints in our scenes requires techniques that process the depth and color vision data efficiently. Considering these difficulties, our primary objective is to generate high-quality textels that can also be rendered on a standard hardware setup using texture mapping. For this work, we have made three contributions. Our first contribution involves the application of a visibility map to efficiently identify visible faces. The second contribution is a technique to reduce ghosting artifacts based on a confidence map. The third contribution yields high-detail textels by adding the mean color and color histogram information to the sigma-outlier detector. The experimental results show that our multi-view texture mapping approach efficiently generates high-quality textels for colored 3D models, while being robust to registration errors.

Paper Nr: 9
Title:

Evaluation of Distance-Aware KinFu Algorithm for Stereo Outdoor Data

Authors:

Hani Javan Hemmat, Egor Bondarev, Gijs Dubbelman and Peter H. N. de With

Abstract: In this paper, we report on experiments on deployment of an extended distance-aware KinFu algorithm, designed to generate 3D model from Kinect data, onto depth frames extracted from stereo camera data. The proposed idea allows to suppress the Kinect usage limitation for outdoor sensing due to the IR interference with sun light. Besides this, exploiting the stereo data enables a hybrid 3D reconstruction system capable of switching between the Kinect depth frames and stereo data depending on the quality and quantity of the 3D and visual features on a scene. While the nature of the stereo sensing and the Kinect depth sensing is completely different, the stereo camera and the Kinect show similar sensitivity to distance capturing. We have evaluated the stereo-based 3D recon- struction with the extended KinFu algorithm with the following distance aware weighting strategies: (a) weight definition to prioritize importance of the sensed data depending on its accuracy, and (b) model updating to decide about the level of influence of the new data on the existing 3D model. The qualitative compari- son of the resulting outdoor 3D models shows higher accuracy and smoothness of models obtained by introduced distance-aware strategies. The quantitative anal- ysis reveals that applying the proposed weighting strategies onto stereo datasets enables to increase robustness of the pose-estimation algorithm and its endurance by factor of two.

Short Papers
Paper Nr: 5
Title:

Vehicle Tracking based on Customized Template Matching

Authors:

Sebastiano Battiato, Giovanni Maria Farinella, Antonino Furnari, Giovanni Puglisi, Anique Snijders and Jelmer Spiekstra

Abstract: In this paper we present a template matching based vehicle tracking algorithm designed for traffic analysis purposes. The proposed approach could be integrated in a system able to understand lane changes, gate passages and other behaviours useful for traffic analysis. After reviewing some state-of-the-art object tracking techniques, the proposed approach is presented as a customization of the template matching algorithm by introducing different modules designed to solve specific issues of the application context. The experiments are performed on a dataset compound by real-world cases of vehicle traffic acquired in different scene contexts (e.g., highway, urban, etc.) and weather conditions (e.g., raining, snowing, etc.). The performances of the proposed approach are compared with respect to a baseline technique based on background-foreground separation.