Merge branch 'master' of https://github.com/Defcon27/Rapid-Image-Seach-using-Parallel-Computing

Author: Defcon27

README.md

@@ -0,0 +1,11 @@
+# CRISPR--CBIR-based-Rapid-Image-Seach-using-Parallel-Computing
+
+#### CRISPR is a CBIR(Content Based Image Search) based Image Search application that is capable of retrieving similar images of query image input from a database of images. This system uses the concept of parallel computing to speed up the search thus reducing the time required to retrieve the images. 
+
+## ABSTRACT
+With the massive  growth of the internet, people can gain access to a massive amount of information. Due to this retrieving the information of interest becomes very difficult. If focused on visual information, the internet contains several kinds of images and other visual information, such as videos, movies in various formats such as JPG, PNG, BMP and even GIF. Hence there is a need for such an image search engine using which the related and exact images can be searched. Image Search Engine that is based on a content-based image retrieval query technique involves providing the CBIR system with a sample image that it will then base its search upon. Content-based image retrieval seeks to find
+methods to index, browse, and query large image databases by using meaningful feature extraction and comparison methods for images. CBIR system uses feature extraction which is the process of obtaining the most relevant information from the original image and represent it in a reduced representation of a set of features like texture, shape using algorithms which process the image data and store them. Feature extraction and representing the processed data from the contents of images is a very challenging task hence various
+feature extraction techniques and algorithms should be applied to use and create different algorithms to increase efficiency in extraction and importantly time consumption of the algorithm. 
+
+But to implement such kind of algorithms in real-world applications we need the algorithm to be executed in the least time possible so as to increase the performance of the system. This speed-up of performance can be achieved since feature extraction and comparison of visual features used for the content-based image retrieval can be realized by using the concept of parallel computing. Since the algorithms dealing with the extraction process are huge and complex, with the help of parallelization this heavy process can be divided into multiple smaller tasks and execute them at the same time. The main purpose of parallelization is to provide simultaneous execution of two or more parts of the program to utilize the CPU resources to the maximum increasing CPU utilization. This helps the program to run faster, smoother, and much efficient in resource utilization. Thus, the
+implementation of parallelization in image search could greatly reduce the retrieval time and improve the performance of retrieval system which is critical in any search applications. Applications of feature extraction from images are limitless. This data can be used in classification, recognition of images in a huge database centers where processing needs to be fast and efficient. Their significant applications in security systems as it is the basis in biometric systems. Since users usually work with a huge number of images, it is important to achieve the highest performance possible from that code. To achieve this, we make use of parallelization.