Resource limitation is one of the challenges in wireless sensor networks (W) that must be considered in developing algorithms and design of networks. Energy supplies, computational power, channel bandwidth and size of memory are among those limitations. Since batteries are the most commonly used source of energy of a node in W, energy usage has influence in the network lifetime. In this dissertation we review the challenges in the design of W. Then, we focus on one of the most important applications, which is surveillance and define two problems related to this application. The main goal of both problems is to select the minimum number of cameras in order to prolong the network lifetime. In the first problem we investigate area coverage with homogeneous resolution in which all the points of the area have the same importance for being covered. To solve the problem we propose three greedy and two evolutionary algorithms. In the second problem target points coverage with heterogeneous resolution is considered. We first introduce this problem as a discrete optimization problem and then prove that this problem is NP-Hard. Then, we propose two central and greedy algorithms in addition to a distributed one to solve the problem. For the specific form of the problem in which we can consider all target points on a line, we propose an optimum solution with low complexity based on trellis structure. Moreover, we examine a problem in general to see if the proposed optimum solution is applicable. In order to do so, we propose a method based on a tree structure. After selecting cameras, if selected nodes want to save captured videos or transfer them, a high volume of memory or a high bandwidth channel are required. On the other hand, sensor networks have limitations in both bandwidth and the size of memory. In order to reduce the size of data, video streams should be coded. In this dissertation we focus on the motion estimation part of a standard codec which is computationally expensive. In order to reduce the complexity burden of this part we propose a simple and regular algorithm. The proposed algorithm is designed based on statistical analyses performed on a number of video sequences. The proposed method is compared with some of the existing methods in terms of R and running time. Keywords Camera Sensor Networks, area coverage, target coverage, homogeneous coverage, heterogeneous coverage, video compression.