Image ROI Area Automatic Extraction And Coding

With the development of multimedia technology and the need of intelligent of information communication, recently the technology of ROI extraction and coding become the hot topic. The technology of ROI extraction is the basic of image content different coding and intelligent video analysis,what’s more,it’s useful for image retrieval、intelligent video analysis、remote medical image diagnosis.ROI Coding technology is used in the areas where the bandwith is limited.The technology of image ROI extraction and coding have following a few questions:Firstly in the pace of image preprocessing,the noise and uneven illumination have great effect on the accuracy rating of ROI extraction.Secondly how to extract the image ROI areas is the main topic in this paper.In this paper, facial feature is regard as the ROI areas and be used in security and protection monitoring.In the end studying the methods of image ROI coding technology is very important in the paper.Excellent image preprocessing method is very important for image ROI extraction.At first this paper studys the methods of noise removal and illumination compensation and uses the improved median filter method and the improved Retinex method to process the images.The experiment proves that this improved method is effective.Image ROI extraction is the important part in this paper.This paper studies some common ROI extraction ways.Thinking about the features of security and protection monitoring the face is regard as ROI areas. After studying the common face detection methods,the improved Adaboost combined with skin color feature method is used in this paper.The experiment proves that this improved method is valid.Nowadays in the basis of human eyes visual characteristics the new ROI extraction methods become the new trend.This paper studies the old common human eys visual ROI extraction methods and comes up with a new method which can improves the precision rate of ROI extraction.The experiment proves that this improved method is in force.After extracting the ROI areas,the ROI coding images methods are used in processing thses images.This paper studies the recent the ways of coding the images in the JPEG2000image coding standard and compare these method’s merit and demerit.This paper puts forward the new method which combines coeffient shifting with rate distortion optimization methods.The experiment proves that this improved method is useful.