**1.2 Multimedia sensing subsystem**

One of the main differences between multimedia sensor networks and other types of sensor networks lies in the nature of how the image sensors perceive information from the environment. Most scalar sensors provide measurements as 1-dimensional data signals. However, image sensors are composed of a large number of photosensitive cells. One measurement of the image sensor provides a 2-dimensional set of data points, which we see as an image. The additional dimensionality of the data set results in richer information content as well as in a higher complexity of data processing and analysis. In addition, a camera's sensing model is inherently different from the sensing model of any other type of sensor. Typically, a scalar sensor collects data from its vicinity, as determined by its sensing range. Multimedia nodes are characterized by a directional sensing model, called Field of View (FoV, see Figure 1), and can capture images of distant/vicinal objects/scenes within its FoV from a certain direction. The object covered by the camera can be distant from the camera and the captured images will depend on the relative positions and orientation of the cameras towards the observed object (Soro & Heinzelman, 2005; Tezcan & Wang, 2008; Adriaens et al., 2006). Because of non-coincidence between neighborhood and sensed region by multimedia nodes, coverage-based techniques in WSN do not satisfy WMSN requirements.

Accordingly, the amount of power consumed in the sensing subsystem of a multimedia sensor node is considerably more than of a scalar ordinary sensor. For example, a

consumption of the rest of the sensor node (Alippi et al., 2007). This can be due to many

 Power hungry transducers. Some sensors intrinsically require high power resources to perform their sampling task. For example, sensing arrays such as CCDs or multimedia sensors (Akyildiz et al., 2007) such as CMOS image sensors generally require a lot of power. Also chemical or biological sensors (Diamond, 2006) can be power hungry as

 Long acquisition time. The acquisition time may be in the order of hundreds of milliseconds or even seconds, especially in the case of multimedia sensors. Hence the energy consumed by the sensing subsystem may be high, even if the sensor power consumption is moderate. In this case reducing communications may be not enough, but energy conservation schemes have to actually reduce the number of acquisitions (i.e. data samples). It should also be pointed out that energy-efficient data acquisition techniques are not exclusively aimed at reducing the energy consumption of the sensing subsystem. By reducing the data sampled by source nodes, they decrease the number of communications as well. Actually, many energy-efficient data-acquisition techniques have been conceived for minimizing the radio energy consumption, under the

 Power hungry A/D converters. Sensors like acoustic and seismic transducers generally require high-rate and high-resolution A/D converters. The power consumption of the converters can account for the most significant power consumption of the sensing

One of the main differences between multimedia sensor networks and other types of sensor networks lies in the nature of how the image sensors perceive information from the environment. Most scalar sensors provide measurements as 1-dimensional data signals. However, image sensors are composed of a large number of photosensitive cells. One measurement of the image sensor provides a 2-dimensional set of data points, which we see as an image. The additional dimensionality of the data set results in richer information content as well as in a higher complexity of data processing and analysis. In addition, a camera's sensing model is inherently different from the sensing model of any other type of sensor. Typically, a scalar sensor collects data from its vicinity, as determined by its sensing range. Multimedia nodes are characterized by a directional sensing model, called Field of View (FoV, see Figure 1), and can capture images of distant/vicinal objects/scenes within its FoV from a certain direction. The object covered by the camera can be distant from the camera and the captured images will depend on the relative positions and orientation of the cameras towards the observed object (Soro & Heinzelman, 2005; Tezcan & Wang, 2008; Adriaens et al., 2006). Because of non-coincidence between neighborhood and sensed region by multimedia nodes, coverage-based techniques in

Accordingly, the amount of power consumed in the sensing subsystem of a multimedia sensor node is considerably more than of a scalar ordinary sensor. For example, a

different factors (Raghunathan et al., 2006):

assumption that the sensor consumption is negligible.

subsystem, as in (Schott et al., 2005).

WSN do not satisfy WMSN requirements.

**1.2 Multimedia sensing subsystem** 

well.

temperature sensor (texas instrument, 2011) as a scalar sensor consumes 6μW for sensing the environment. To have a view of multimedia sensors power consumption, table 1 shows the power consumed by four classes of cameras that are available today either as prototypes or as commercial products. At the lowest end of the spectrum is tiny Cyclops (Rahimi et al., 2005) that consumes a mere 46mW and can capture low resolution video. CMU-Cams (Rowe et al., 2002) are cell-phone class cameras with on-board processing for motion detection, histogram computation, etc. At the high-end, web-cams can capture high-resolution video at full frame rate while consuming 200mW, whereas Pan-Tilt-Zoom cameras are re-targetable sensors that produce high quality video while consuming 1W. It is noticeable that the mentioned power amounts are the power consumed by the camera sensors without considering the power consumed by the host motes, see (Tavli et al, 2011) for a survey of visual network platforms.

Fig. 1. The Field of View (FoV) of a multimedia sensor node.


Table 1. Power consumption and capabilities of four classes of camera sensors.

On the other hand, given the large amount of data generated by the multimedia nodes, both processing and transmitting image data are quite costly in terms of energy, much more so than for other types of sensor networks. Furthermore, visual sensor networks require large bandwidth for transmitting image data. Thus both energy and bandwidth are even more constrained than in other types of wireless sensor networks.

In this chapter, we describe a power efficient mechanism for managing the sensing subsystem of multimedia sensor nodes for surveillance in WMSNs. For this purpose, the deployed multimedia nodes are clustered according to their common covering regions and the clusters are managed to schedule the members to collaboratively survey the sensing area in a dutycycled manner. With avoiding acquisition of redundant and correlated data, not only the sensing subsystem of nodes save its energy, but also the transmission and processing subsystems meet an optimized amount of data to be transmitted/processed and thus can conserve their residual energy. Therefore, the network lifetime is considerably prolonged.

The chapter is organized as follows. In section 2 we present an overview of work related to sensor management and scheduling policies. A surveillance mechanism with its details in grouping, management and scheduling multimedia nodes to be energy efficient is explained in section 3. Finally, the future work and conclusions are derived.
