Progressive Image Transmission Using Significant Features and Linear Regression Prediction

Progressive image transmission is often used in low bandwidth communications or in limited bandwidth environments. When a user browses an image on a website, the web server first sends a low quality image to the receiver and then gradually improves the image quality by subsequent transmissions .Because the image size becomes larger and larger, it is essential to develop a flexible and efficient technique to help users quickly browse or retrieve images on a website. For this reason, using significant features along with a prediction technique for progressive image transmission is proposed in this paper. In our method, significant features of an image need to be extracted first. Next, these features, according to their priority, are sent to the receiver. For reducing the encoding cost of the transmitted data, three approaches, sampling, quantization, and linear regression prediction, are also presented.