As a video encoder and a video decoder are adopted in most of battery-operated portable devices, e.g., cell phone, digital camera, laptop, camcorder, etc, low-energy video encoding and video decoding become the critical issues in such system design to extend battery lifetime. To tackle the issues, in this thesis, I present energy reduction methods in a video encoder/decoder by utilizing dynamic voltage and frequency scaling (DVFS) exploiting three outstanding characteristics: i) runtime distribution on both computational workload and memory stall time in a video encoder/decoder, ii) varying parallelism over each path in a parallelized video encoder/decoder running, and iii) energy-rate-distortion tradeoffs. First, in Chapter 1, I propose an online DVFS method which takes into account the dynamism of both program phase behavior and runtime distribution of processor computational workload and memory stall time. The online DVFS problem is addressed in two ways: intra-phase workload prediction and program phase detection. The intra-phase workload prediction is to predict the workload based on the runtime distribution of computational workload and memory stall time in the current program phase in order to over-come the conventional pessimistic assumption on the remaining workload, i.e., worst-case execution cycle. The program phase detection is to identify to which program phase the current instant belongs and then to obtain the predicted workload corresponding to the detected program phase, which is used to set voltage and frequency during the program phase. Then, in Chapter 2, I extend the runtime distribution-aware DVFS method for a multi-processor with judiciously exploiting slack by considering the varying parallelism over each path in a task graph running on a multi-processor. Finally, in Chapter 3, I present a low-energy wireless video sensor node system consisting of event detector, video encoder, transceiver, and memory. To reduce the energy consumption, ...