PENERAPAN ALGORITME KOMPRESI JPEG DAN METODE FUZZY C-MEANS PADA KOMPRESI CITRA BERBASIS ENTROPI

This entropy Advances in technology which rapidly grows demand a lot of information to be processed, stored and delivered. With the amount of the information grows bigger means the needs for memory to represent the information will also grows bigger. A digital image is an information in the form of 2D (two dimension) which is processed through the visual interpretation by human eye. To save more information contained in the image therefore image compression is necessary. JPEG is one of many lossy compression methods used in image compression. JPEG is also widely known and commonly used nowadays. In this final project, the author will use the JPEG's compression algorithm to generate an entropy-based image compression. -based image compression will be tested on image with 2 different types of initial compression format, .jpg and .png. Before sample image is compressed by JPEG, it will be grouped into clusters based on its information with the help of Fuzzy C-Means (FCM). To speed up the iteration of FCM, the author will do an initialization to determine cluster's first centroid. Contained information on each cluster's can be determined based on its entropy, and then JPEG will apply compression with different quality scale on each cluster. Final results obtained is complete image formed from all clusters. Where cluster with low entropy will not be compressed and cluster with high entropy will be compressed. Based on assumption that a cluster with low entropy contains important information and a cluster with high entropy contains less important information. Sample image with .png compression format give out better output than sample image with .jpg compression format, whether in terms size, qualitative-based comparison (visual) and quantitative-based comparison (mathematical calculation)