The 2D-wavelet transform of images is, among others, used for image compression (either for still images or for video sequences). A wavelet transformed image has the special property that the image content is split into a lowpass component and a highpass component. The highpass component usually consists mostly of small coefficients and can be compressed efficiently using an entropy coder. A Quadtree coder [1] is such an entropy coder which achieves good compression rates and in addition has a number of scalability features. A Quadtree coder compresses a wavelet image progressively i.e. it first encodes the most significant digits of the wavelet coefficients allowing for progressive approximations of the image at the decoder side. During decoding, once enough accuracy is achieved the decoding process can be aborted without the need to decode the rest of the encoded video stream. A second property of the Quadtree algorithm is that it reorders the content of the coded bit stream so that information that has a larger contribution to the overall image quality is shifted towards the start of the video stream. In this article we present results regarding a hardware implementation of the Quadtree algorithm. We find that the progressive coding and the reordering of the video stream during Quadtree coding requires a tailored approach to its hardware implementation in order to achieve an efficient memory use and a high compression speed. We present a number of techniques to deal with these challenges and give simulation results of the expected performance of the Quadtree coder.