Performance evaluation of crowd image analysis using the PETS2009 dataset

This paper presents the PETS2009 outdoor crowd image analysis surveillance dataset and the performance evaluation of people counting, detection and tracking results using the dataset submitted to five IEEE Performance Evaluation of Tracking and Surveillance (PETS) workshops. The evaluation was carried out using well established metrics developed in the Video Analysis and Content Extraction (VACE) programme and the CLassification of Events, Activities, and Relationships (CLEAR) consortium. The comparative evaluation highlights the detection and tracking performance of the authors’ systems in areas such as precision, accuracy and robustness and provides a brief analysis of the metrics themselves to provide further insights into the performance of the authors’ systems.

Overall evaluation of the PETS2009 results

This paper presents the results of the crowd image analysis challenge, as part of the PETS 2009 workshop. The evaluation is carried out using a selection of the metrics available in the Video Analysis and Content Extraction (VACE) program and the CLassification of Events, Activities, and Relationships (CLEAR) consortium. The evaluation highlights the strengths of the authors’ systems in areas such as precision, accuracy and robustness.

PETS2009 and Winter-PETS 2009 results: a combined evaluation

This paper presents the results of the crowd image analysis challenge of the Winter PETS 2009 workshop. The evaluation is carried out using a selection of the metrics developed in the Video Analysis and Content Extraction (VACE) program and the CLassification of Events, Activities, and Relationships (CLEAR) consortium [13]. The evaluation highlights the detection and tracking performance of the authors’systems in areas such as precision, accuracy and robustness. The performance is also compared to the PETS 2009 submitted results.

PETS2010 and PETS2009 evaluation of results using individual ground truthed single views

This paper presents the results of the crowd image analysis challenge of the PETS2010 workshop. The evaluation was carried out using a selection of the metrics developed in the Video Analysis and Content Extraction (VACE) program and the CLassification of Events, Activities, and Relationships (CLEAR) consortium. The PETS 2010 evaluation was performed using new ground truthing create from each independant two dimensional view. In addition, the performance of the submissions to the PETS 2009 and Winter-PETS 2009 were evaluated and included in the results. The evaluation highlights the detection and tracking performance of the authors’ systems in areas such as precision, accuracy and robustness.

Supplementary feeding of wild birds indirectly affects ground beetle populations in suburban gardens

Supplementary feeding of wild birds by domestic garden-holders is a globally widespread and popular form of human–wildlife interaction, particularly in urban areas. Vast amounts of energy are thus being added to garden ecosystems. However, the potential indirect effects of this activity on non-avian species have been little studied to date, with the only two previous studies taking place under experimentally manipulated conditions. Here we present the first evidence of a localised depletive effect of wild bird feeding on ground beetles (Coleoptera: Carabidae) in suburban gardens under the usual feeding patterns of the garden-holders. We trapped significantly fewer ground beetles directly under bird-feeding stations than in matched areas of habitat away from feeders. Video analysis also revealed significantly higher activity by ground-foraging birds under the feeding stations than in the control areas. Small mammal trapping revealed no evidence that these species differ in abundance between gardens with and without bird feeders. We therefore suggest that local increases in ground-foraging activity by bird species whose diets encompass arthropods as well as seed material are responsible for the reduction in ground beetle numbers. Our work therefore illustrates that providing food for wild birds can have indirect negative effects on palatable prey species under typical conditions.

PETS 2015: datasets and challenge

This paper presents the two datasets (ARENA and P5) and the challenge that form a part of the PETS 2015 workshop. The datasets consist of scenarios recorded by us- ing multiple visual and thermal sensors. The scenarios in ARENA dataset involve different staged activities around a parked vehicle in a parking lot in UK and those in P5 dataset involve different staged activities around the perimeter of a nuclear power plant in Sweden. The scenarios of each dataset are grouped into ‘Normal’, ‘Warning’ and ‘Alarm’ categories. The Challenge specifically includes tasks that account for different steps in a video understanding system: Low-Level Video Analysis (object detection and tracking), Mid-Level Video Analysis (‘atomic’ event detection) and High-Level Video Analysis (‘complex’ event detection). The evaluation methodology used for the Challenge includes well-established measures.

PETS 2014: dataset and challenge

This paper describes the dataset and vision challenges that form part of the PETS 2014 workshop. The datasets are multisensor sequences containing different activities around a parked vehicle in a parking lot. The dataset scenarios were filmed from multiple cameras mounted on the vehicle itself and involve multiple actors. In PETS2014 workshop, 22 acted scenarios are provided of abnormal behaviour around the parked vehicle. The aim in PETS 2014 is to provide a standard benchmark that indicates how detection, tracking, abnormality and behaviour analysis systems perform against a common database. The dataset specifically addresses several vision challenges corresponding to different steps in a video understanding system: Low-Level Video Analysis (object detection and tracking), Mid-Level Video Analysis (‘simple’ event detection: the behaviour recognition of a single actor) and High-Level Video Analysis (‘complex’ event detection: the behaviour and interaction recognition of several actors).

Meeting detection in video through semantic analysis

In this paper we present a novel approach to detect people meeting. The proposed approach works by translating people behaviour from trajectory information into semantic terms. Having available a semantic model of the meeting behaviour, the event detection is performed in the semantic domain. The model is learnt employing a soft-computing clustering algorithm that combines trajectory information and motion semantic terms. A stable representation can be obtained from a series of examples. Results obtained on a series of videos with different types of meeting situations show that the proposed approach can learn a generic model that can effectively be applied on the behaviour recognition of meeting situations.

Innovative use of vodcast (Video-Podcast) to enrich learning experience in structures laboratory

Students may have difficulty in understanding some of the complex concepts which they have been taught in the general areas of science and engineering. Whilst practical work such as a laboratory based examination of the performance of structures has an important role in knowledge construction this does have some limitations. Blended learning supports different learning styles, hence further benefits knowledge building. This research involves an empirical study of how vodcasts (video-podcasts) can be used to enrich learning experience in the structural properties of materials laboratory of an undergraduate course. Students were given the opportunity of downloading and viewing the vodcasts on the theory before and after the experimental work. It is the choice of the students when (before or after, before and after) and how many times they would like to view the vodcasts. In blended learning, the combination of face-to-face teaching, vodcasts, printed materials, practical experiments, writing reports and instructors’ feedbacks benefits different learning styles of the learners. For the preparation of the practical, the students were informed about the availability of the vodcasts prior to the practical session. After the practical work, students submitted an individual laboratory report for the assessment of the structures laboratory. The data collection consisted of a questionnaire completed by the students, follow-up semi-structured interviews and the practical reports submitted by them for assessment. The results from the questionnaire were analysed quantitatively, whilst the data from the assessment reports were analysed qualitatively. The analysis shows that most of the students who have not fully grasped the theory after the practical, managed to gain the required knowledge by viewing the vodcasts. According to their feedbacks, the students felt that they have control over how to use the material and to view it as many times as they wish. Some students who have understood the theory may choose to view it once or not at all. Their understanding was demonstrated by their explanations in their reports, and was illustrated by the approach they took to explicate the results of their experimental work. The research findings are valuable to instructors who design, develop and deliver different types of blended learning, and are beneficial to learners who try different blended approaches. Recommendations were made on the role of the innovative application of vodcasts in the knowledge construction for structures laboratory and to guide future work in this area of research.

Blended learning approach to include the innovative use of vodcasts (video-podcasts) to enrich students learning experience in structures laboratory

Students may have difficulty in understanding some of the complex concepts which they have been taught in the general areas of science and engineering. Whilst practical work such as a laboratory based examination of the performance of structures has an important role in knowledge construction this does have some limitations. Blended learning supports different learning styles, hence further benefits knowledge building. This research involves the empirical studies of how an innovative use of vodcasts (video-podcasts) can enrich learning experience in the structural properties of materials laboratory of an undergraduate course. Students were given the opportunity of downloading and viewing the vodcasts on the theory before and after the experimental work. It is the choice of the students when (before or after, before and after) and how many times they would like to view the vodcasts. In blended learning, the combination of face-to-face teaching, vodcasts, printed materials, practical experiments, writing reports and instructors’ feedbacks benefits different learning styles of the learners. For the preparation of the practical laboratory work, the students were informed about the availability of the vodcasts prior to the practical session. After the practical work, students submit an individual laboratory report for the assessment of the structures laboratory. The data collection consists of a questionnaire completed by the students, and the practical reports submitted by them for assessment. The results from the questionnaire were analysed quantitatively, whilst the data from the assessment reports were analysed qualitatively. The analysis shows that students who have not fully grasped the theory after the practical were successful in gaining the required knowledge by viewing the vodcasts. Some students who have understood the theory may choose to view it once or not at all. Their understanding was demonstrated by the quality of their explanations in their reports. This is illustrated by the approach they took to explicate the results of their experimental work, for example, they can explain how to calculate the Young’s Modulus properly and provided the correct value for it. The research findings are valuable to instructors who design, develop and deliver different types of blended learning, and beneficial to learners who try different blended approaches. Recommendations were made on the role of the innovative application of vodcasts in the knowledge construction for structures laboratory and to guide future work in this area of research.

An FLIR video surveillance system to avoid bridge-ship collision

In this paper, a forward-looking infrared (FLIR) video surveillance system is presented for collision avoidance of moving ships to bridge piers. An image pre-processing algorithm is proposed to reduce clutter noises by multi-scale fractal analysis, in which the blanket method is used for fractal feature computation. Then, the moving ship detection algorithm is developed from image differentials of the fractal feature in the region of surveillance between regularly interval frames. Experimental results have shown that the approach is feasible and effective. It has achieved real-time and reliable alert to avoid collisions of moving ships to bridge piers

Semi-automatic knowledge extraction, representation and context-sensitive intelligent retrieval of video content using collateral context modelling with scalable ontological networks

Automatic indexing and retrieval of digital data poses major challenges. The main problem arises from the ever increasing mass of digital media and the lack of efficient methods for indexing and retrieval of such data based on the semantic content rather than keywords. To enable intelligent web interactions, or even web filtering, we need to be capable of interpreting the information base in an intelligent manner. For a number of years research has been ongoing in the field of ontological engineering with the aim of using ontologies to add such (meta) knowledge to information. In this paper, we describe the architecture of a system (Dynamic REtrieval Analysis and semantic metadata Management (DREAM)) designed to automatically and intelligently index huge repositories of special effects video clips, based on their semantic content, using a network of scalable ontologies to enable intelligent retrieval. The DREAM Demonstrator has been evaluated as deployed in the film post-production phase to support the process of storage, indexing and retrieval of large data sets of special effects video clips as an exemplar application domain. This paper provides its performance and usability results and highlights the scope for future enhancements of the DREAM architecture which has proven successful in its first and possibly most challenging proving ground, namely film production, where it is already in routine use within our test bed Partners' creative processes. (C) 2009 Published by Elsevier B.V.

An analysis on the performance of the forward and return interaction channels in DVB-RCS

The next generation consumer level interactive services require reliable and constant communication for both mobile and static users. The Digital Video Broadcasting ( DVB) group has exploited the rapidly increasing satellite technology for the provision of interactive services and launched a standard called Digital Video Broadcast through Return Channel Satellite (DYB-RCS). DVB-RCS relies on DVB-Satellite (DVB-S) for the provision of forward channel. The Digital Signal processing (DSP) implemented in the satellite channel adapter block of these standards use powerful channel coding and modulation techniques. The investigation is concentrated towards the Forward Error Correction (FEC) of the satellite channel adapter block, which will help in determining, how the technology copes with the varying channel conditions and user requirements(1).

Video activity extraction and reporting with incremental unsupervised learning

The present work presents a new method for activity extraction and reporting from video based on the aggregation of fuzzy relations. Trajectory clustering is first employed mainly to discover the points of entry and exit of mobiles appearing in the scene. In a second step, proximity relations between resulting clusters of detected mobiles and contextual elements from the scene are modeled employing fuzzy relations. These can then be aggregated employing typical soft-computing algebra. A clustering algorithm based on the transitive closure calculation of the fuzzy relations allows building the structure of the scene and characterises the ongoing different activities of the scene. Discovered activity zones can be reported as activity maps with different granularities thanks to the analysis of the transitive closure matrix. Taking advantage of the soft relation properties, activity zones and related activities can be labeled in a more human-like language. We present results obtained on real videos corresponding to apron monitoring in the Toulouse airport in France.

Analysis of a DVB-T compliant receiver simulation under various multipath conditions

Since 1966, coded orthogonal frequency division multiplexing (COFDM) has been investigated to determine the possibility of reducing the overall throughput of a digitally modulated terrestrial television channel. In the investigations, many assumptions have emerged. One common misconception is that in a terrestrial environment, COFDM has an inherent immunity to multipath interference. A theoretical analysis of a multipath channel, along with simulation results has shown that this assumption does not hold the information is considered when including the radio frequency modulation and demodulation. This paper presents a background into the inception of COFDM, a mathematical analysis of the digitally modulated television signal under multipath conditions and the results of a European Digital Video Broadcasting-Terrestrial (DVB-T) compliant simulation model with MPEG-2 bitstreams transmitted under various multipath conditions.