International Journal of Computer Applications (0975 – 8887) Volume 78 – No.10, September 2013 11 Robust and Efficient ‘RGB’ based Fractal Image Compression: Flower Pollination based Optimization Gaganpreet Kaur Dheerendra Singh, PhD Manjinder Kaur Assistant Professor Professor & Head Student Dept. of CSE, SGGSWU SUSCET Tangori Dept. of CSE, SGGSWU Fatehgarh Sahib, Punjab Mohali, Punjab Fatehgarh Sahib, Punjab ABSTRACT Fractal image compression uses the property of self- similarity in an image and utilizes the partitioned iterated function system to encode it. Fractal image compression is attractive because of high compression ratio, fast decompression and multi-resolution properties. The main drawback of Fractal Image Compression is the high computational cost and is the poor retrieved image qualities. To overcome this drawback, we design a new algorithm which is based on Pollination Based Optimization which is used to classify the phantom, satellite and rural image dataset. Flower Pollination Based Optimization is nature inspired algorithm which decreases the search complexity of matching between range block and domain block. Also, the optimization technique has effectively reduced the encoding time while retaining the quality of the image. Peak signal to noise ratio, entropy, compression ratio and mean square error is found for phantom, rural and satellite images data set. This new method showed improved highly accurate results. General terms Optimization, Soft Computing Keywords Pollination Based Optimization (PBO), Fractal Image Compression (FIC), Satellite Image ,Phantom Images, Partitioned Iterated Function System, Range Block, Domain Block, Peak Signal Noise Ratio (PSNR), Mean Square Error (MSE), Compression Ratio, Entropy. 1. INTRODUCTION The foundation of fractal image goes to Barnsley & Jacquin. In fractal image compression an image is represented by fractals rather than pixel. Block segmentation, region segmentation and the cross searching method lie under the umbrella of self- similarity algorithms. A fractal is worked as Iterated Function System which is a group of affine transformations. The main work for fractal coding is to extract the fractals which are used for approximation to the original image. These fractals are represented as a set of affine transformations. It has high compression ratio and simple decompression method. The main drawback is larger computational time for image compression. In order to reduce the computation time different optimization techniques have been proposed. It is found that the computational time can be improved by performing the search in a small sub-set of domain pool rather than over the whole space. In fractal image compression, the image is divided into a number of block domains with arbitrary size. Then, the image is divided again into block ranges with size less than that of the block domain. The main objective of this paper is to develop an efficient optimization technique for fractal image compression which involves classifying the domain pool blocks for a gray level image and color image, thus improving the encoding time and quality of images. In this paper, flower Pollination Based Optimization approach is used for fractal image compression. The paper is organized as follows. In section 2, pollination based Optimization (PBO) is discussed. In section 3, the proposed methodology is discussed. Experimental results on images are presented in section 4. Finally, in Section 5, some conclusions and directions for future work are discussed. 2. POLLINATION BASED OPTIMIZATION Pollination is a process of transfer of pollen from the male parts of a flower called anther to the female part called stigma of a flower. Pollination is of two types: Biotic and Abiotic pollination. In biotic pollination pollen is transferred using insects and animals while in abiotic pollination more number of pollinators is required like wind etc. Plants have both male as well as female organs. The transfer of pollen from one plant to other is considered to be good. The floral display, fragrance and nectar attract pollinators, which leads to pollination. The pollination in plants is 6 week programme. The pollination model used was suggested by Thakar et al. [4] and later modified by Prajakta V. Belsare et. al. [5]. The reproductive success for every plant can be modeled by the following expression:                 …. (1) This was further implemented for application of battery charger to generate rule base using fuzzy by Kumar et al. [2]. 3. PROPOSED METHODOLOGY & IMPLEMENTATION In this work mean square error is used as fitness function. The image as a whole is taken rather than taking single pixel in image. Image is subdivided into R, G, and B components. Pollination parameters like average display, nectar content etc. are initialized.