Daichi Shimada

Image Compression for Remote Desktop for Engineering Cloud

With the development of the cloud computing, many services that provide a desktop environment over a network are coming into widespread use. In recent years, the demand that globalized product developments and cost reductions is increasing. This leads to a growing demand for using these services in product design and development activities, a challenge in this situation is to make 3D-CAD and CAE software available over remote desktop services of the cloud, namely Engineering Cloud. In this paper, we propose a lossy image compression method for 3D-CAD and CAE software at high compression ratio. This method extracts constant gradients by a frequency transform, which exploits the nature as artificial images in that the local variations in pixel value are constant. We demonstrate that this method achieves a 1.4 times improvement in compression ratio as compared with conventional JPEG. We also apply this method to a remote desktop system, which demonstrates that the bandwidth is reduced by 43% of JPEG case.