Visual Human Tracking in Wireless Cameras Networks A SURF-Based Approach

In wireless camera networks, the communication load between cameras is a major concern for visual tracking. To save the bandwidth, traditional applications transfer the spatial coordinates under the precondition of camera calibration, which is computationally unreasonable for large and mobile camera networks. In this chapter, we exploit the use of distinctive and fast to compute local features to represent the non-rigid targets. Transmission of feature descriptors between cameras is done without any calibration. Combining the haar-like patterns and relative color information, our local features succeed to re-identify and relocate the target among the distributed cameras. Furthermore, efficient interest point detection and matching scheme are proposed for the visual tracking under real-time constraints.

Low Level Representation of Data for Visual Sensor Network

Wireless Sensor Network future direction is going towards more complex sensor such as camera sensor. Therefore, a very active research field is Visual Sensor Network. This type of network brings new challenges such as processing and transmitting a massive amount of data generated by the camera sensor. The efforts into decreasing the amount of data to be transmitted are going towards two directions: data encoding and data filtering. This chapter introduces an algorithm for each direction. Visual data encoding is performed by means of Predictive Video Encoding using Phase-Only Correlation function to achieve motion estimation. Visual data filtering is done at the lowest level of abstraction and is performed in three phases: pixel classification, background update and detection. The algorithms involved in each phase are light in terms of complexity and memory resources.

Applications for Wireless Visual Sensor Networks

In this chapter we will describe our work to set up a large scale wireless visual sensor network in a Swedish zoo. It is located close to the Arctic Circle making the environment very hard for this type of deployment. The goal is to make the zoo digitally enhanced, leading to a more attractive and interactive zoo. To reach this goal the sensed data will be processed and semantic information will be used to support interaction design, which is a key component to provide a new type of experience for the visitors. In this chapter we will describe our research work related to the various aspects of a digital zoo.

Energy Efficiency and Reliability Considerations of a Compressive Sensing Technique in Wireless Visual Sensor Networks

This chapter focuses on the energy efficiency and reliability issues when applying the novel compressive sensing technique in wireless visual sensor networks. An explanation is given for why compressive sensing is useful for visual sensor networks. The relationships between sparsity control and compression ratio, the effect of block-based sampling on reconstruction quality, complexity consideration of reconstruction process for real-time applications, and compensation for packets missing in network flows are discussed. We analyse the effectiveness of using the 2-dimensional Haar wavelet transform for sparsity control, the difference between compressive sampling in spatial and frequency domains, and the computation of the prime-dual optimisation method and the log barrier algorithm for reconstruction. The effectiveness of the approach on recovered image quality is evaluated using Euclidean distance and variance analysis.

Quantized Variational Filtering for Bayesian Inference in Wireless Sensor Networks

Simulation results demonstrate the significantly improved performance of our approach.

Visual Sensor Network Technology and Its Applications

The combination of image sensors with wireless sensor network (WSN) technology has resulted in a new network technology called a visual sensor network (VSN). On the one hand, VSNs can be seen as an extension of traditional WSNs where image sensors have replaced scalar sensors. On the other hand, the use of image sensors in VSNs brings with it a different set of practical and research challenges. This is because image sensors generate a very high amount of data that would have to be processed and transmitted within the network. In this chapter, we present an introduction to VSN technology and provide an overview of research issues and trends. Issues related to energy efficient processing, collaborative processing, and hardware technology will be highlighted. This chapter will also give a brief introduction to the other chapters in the book with a focus on showing how the topics covered in each chapter relate to the overall picture of visual information processing in wireless sensor network environments.

Object Association through Multiple Camera Collaboration for Large-Scale Surveillance System

In this chapter, we present an object association method through multiple camera collaboration for a large-scale surveillance system. The object association is achieved by locally generating homographic lines on targets in collaborating cameras. In order to maintain the object association with the insufficient separation between homographic lines due to densely populated objects, homographic points are generated in 3-D with estimated heights. The heights of targets are estimated by the linear least-squares using normal equations. The object association is confirmed by finding the pairs of the correspondences minimizing the distance between them. The proposed method is verified with real video sequences. The simulation result demonstrates that the proposed method is robust against false association because it considers all the possible pairing cases of occluded targets.

Wireless Video Sensor Networks

Wireless Video Sensors Networks (WVSNs) generally suffer from the constraint that their sensor nodes must consume very little power. In this rapidly emerging video application, the traditional video coding architecture cannot be used due to its high encoding complexity. Thankfully, some theorems from Information Theory suggest that this problem can be solved by shifting the encoder tasks, partially or totally, to the decoder. These theorems are employed in the design of so-called Distributed Video Coding (DVC) solutions, the subject matter of this chapter. The chapter not only introduces the DVC but also reviews some important developments of the popular Stanford Wyner-Ziv coding architecture and caps it with latest research trends highlighting a Region-Based-Wyner-Ziv video codec that enables low-complexity encoding while achieving high compression efficiency.

High-Level Information Fusion in Visual Sensor Networks

Information fusion techniques combine data from multiple sensors, along with additional information and knowledge, to obtain better estimates of the observed scenario than could be achieved by the use of single sensors or information sources alone. According to the JDL fusion process model, high-level information fusion is concerned with the computation of a scene representation in terms of abstract entities such as activities and threats, as well as estimating the relationships among these entities. Recent experiences confirm that context knowledge plays a key role in the new-generation high-level fusion systems, especially in those involving complex scenarios that cause the failure of classical statistical techniques –as it happens in visual sensor networks. In this chapter, we study the architectural and functional issues of applying context information to improve high-level fusion procedures, with a particular focus on visual data applications. The use of formal knowledge representations (e.g. ontologies) is a promising advance in this direction, but there are still some unresolved questions that must be more extensively researched.

Distributed Visual Surveillance with Resource Constrained Embedded Systems

In this chapter, we present an innovative technique of line sensor based image capturing and processing in order to detect moving objects such as vehicles. Line Sensor techniques, when used in MWSN, may achieve faster processing results with much less storage and bandwidth requirements while conserving node energy. Line Sensor based processing algorithms provide novel ways for object counting, classification and speed measurement. This solution presents itself as an ideal low-cost candidate for Intelligent Transport Systems (ITS) to monitor and control urban traffic.