Research And Application Of Image Compression Technology Based Seam Carving

21st century information age has been sounded for many years, now it entered a mature stage. The number of sites are also increasing fast, in order to stand out in more and more websites, the only way is optimizing the site itself. As the information giant Google said: "When a page loads longer than 1.5 seconds, the search engine will reduce the site’s search ranking(Page Rank)". That user experience is proportional to page load speed, and web users are getting more and more impatient. Avidan and Shamir proposed Seam Carving technology can provide technical support to optimize the picture for website optimization, to make web pages load faster, and then become more competitive on the market. Image scaling is an important part of digital image processing technology, the traditional methods such as image scaling linearly scaling method(Scale), clipping method(Crop), etc. these methods have some serious defects such as image content deformity and missing its content.The Seam Carving studied in this paper is based on a picture image content scaling mode, you can reduce the size of the picture in the same time maximize the retention of important content of the picture, so that enhance the user experience, the sites are more slimming, website optimization degrees more high. Based on Seam Carving technology author presents an image compression method. The first step compute neighboring pixels’ RGB values difference,then squaring and adding to get the energy value of the point; the second step calculates the energy value of the picture to get the whole energy matrix of pixels, wherein the energy value of a point higher represents it is the cut-off point to other regions, the point is an important element of the film; the third step is by constructing a directed graph form, use the shortest path algorithm or dynamic programming algorithm to find the minimum sum of the energy line in horizontal and vertical direction;the fourth step is to delete these low-energy line in the picture and moved forward the lines that after these deleted lines one cell up to keep the picture quad-shape. Finally,we can achieve the effect of the compressed image, it can better avoid the picture important information missing.Experimental results show that, the intelligent selective deletion energy line algorithm can realized Seam Carving algorithms to achieve better image when zoomed images retain maximum content than normal crop or Proportional compression algorithm. As far as image optimization concerned, the outcome can improve the site picture interactivity, friendliness, and the size of the site has a direct impact on website page loads time. So this is just proof of Seam Carving plays a important role on the optimization of the website.