In this paper, we present a two layer, scalable video cod-A two layer scalable video coding scheme based on a spatiotemporal pyramid is presented. The key component is a novel, effective spatio-temporal video interpolator which is proposed to yield the best prediction from the reduced resolution layer to the higher resolution layer. The proposed video interpolator gives visually correct predicted images and minimizes spatial and the temporal prediction artifacts such as aliasing, uncovered background and blocking effects. Independent bit allocation, which is one of the main requirements of scalable coding, is provided by decoupling of the layers through multistage quantization. The bit rate for the lower resolution channel is directly assigned to the lower resolution layer. However, the rate for the higher resolution channel is distributed between upgrading the lower resolution layer and the higher resolution layer. The decoupled layers are encoded by an MPEG like coder. The described scheme is for a two-layer coder but it is easily extendable to pyramids with more layers.ing scheme based on a spatial pyramid. A spatio-teInpora1 video interpolator is proposed for optimum prediction of the higher resolution layers. The video interpolator interactively combines the spatio-temporal information in a spatially varying way. It reduces artifacts, such as uncovered background, blocking effects, and aliasing by selecting the least distorted regions from each of the temporally predicted and/or spatially interpolated images. Although spatial scalability infers dependencies between the layers, compatible coding requires independent bit allocation for each channel, i.e the rate allocated to the lower resolution layer should not affect the performance of the higher resolution layer. Independent bit allocation is provided by decoupling the layers to some extent by multistage quantization, which is applied to the lower resolution layer.The paper is organized as follows. Section 2 describes the basic unit of the scalable video coder. Section 3 presents a detailed description of the proposed video interpolator. Section 4 provides the scalable coding bit allocation based