Certain investigations on content based image retrieval using machine learning techniques

Abstract

Image processing plays a significant role in evaluating images with respect to numerous criterions. Enhancement, Restoration and compression of images are the three different types of image processing practices which reduce the storage space complexity. In image processing technique, segmentation is an important step which performs a major role throughout the process. Image Segmentation based enhancement technique is extensively used to evaluate the medical images. Generally an image is formed by grouping various numbers of pixels. Depending on the pixel intensities available inside the image, distinct regions of an image can be kept apart. Edge and boundary based segmentations are the two classifications of segmentation procedures. Among the two segmentation procedure, boundary based segmentation is used widely. The various applications of image segmentation are: content-based image retrieval, Medical imaging, object identification and recognition task, automated traffic monitoring system and Video based surveillance system, etc. Within the sight of a few medical imaging techniques, Magnetic Resonance Imaging (MRI) technique holds a prevalent position in the region of recognizing the organs of human body. Inside the human body cancer cells play a significant role in tumour formation. It is identified that nearly 3 illion people surviving in India is affected by cancer, out of which 1 million people are diagnosed with new forms of cancer in the survey taken by Times of India . It leads to the formation and development of lesions and tumours inside the human body and the consequences created by tumour are highly pervasive. A clear experimentation and learning are required by a radiologist to recognize the pathologies embedded within the tissue of brain whereas tissues in brain are the most convoluted part of human body. newline