Content image Retrieval Based on using open Computer Vision and Deep Learning Techniques

Abstract

Content Based Image Retrieval is one of the key Information Retrieval sub field of Image Processing In research. The aim of CBIR Technique is to recognize lexically similar images in reference to a query image specified by the end user. Due to With ever-increasing image collections from multiple sources including companies, military and civilians, browsing and searching across image archives is essential. Content Based Image Retrieval (CBIR) involves multiple techniques to image processing by evaluating image content, such as color, pattern, structure and salient features. By using Filtering technique we can enhance or change the part of input image data. Using edge detection, various type of contents present in image can be retrieved. In deep learning, the model is prepared and it will train on the set of pre-defined images. The accuracy of retrieving content of the image depends upon the model and its training and Various resources generating big data are cc camas, social media, cloud data, IoT. In the proposed approach to perform image classification and to process the image using Filtering and Edge Detection of Artificial Intelligence, Deep learning Techniques [CNN].  Such like Classification using SVM classifier for learning and recognition.