Thesis in Image Processing
Image Processing
Processing image using any signal processing form taking image or video as input and the output in the form of image or the specifications of image is called image processing. The processes that resemble the process of image processing includes computer graphics and computer vision.
There are currently three types of image processing: digital image processing, analog image processing and optical image processing. With the upsurge in scientific visualization, image processing is nowadays trending as an area of interest for scholars and researchers.
Image processing is among fastest growing technologies and it also has applications in the field of business. Image processing can also be used for security purposes as data can be encrypted in the image and can be sent from one place to other. There are wide areas of research in image processing like: cancer imaging, Brain Imaging, Image processing, Imaging Technology, Development of automated software, Development of instrumentation.
To improve the results of image processing parallel and distributed computing paradigms are employed. The technique of image processing includes extracting the parameters and features of image and applying algorithms as per the requirements of one’s research.
Image processing is one of the fastest growing areas in the field of research and development and is also of great interest for researchers. The areas of research in image processing includes watermarking, steganography, fusion, face recognition, quality recognition, segmentation, enhancement, noising and denoising, edge detection, character recognition, currency detection, image registration etc. the analyst having interest in image processing can select any area for carrying out their research.
Image Processing Techniques for Thesis Implementation
-
It is the process of automatically identifying a person from a digital image or a video .This method is based on comparison of the selected facial feature of an image .some features like ,,nose shape, skin color ,face curves, position of face, cheek bone ,jaw etc are selected and are store or save ,whenever the caparison is to be made these selected features are compared with the person in order to check whether the person is same or not .it is basically used for the security purpose . Facial recognition can be used not just to identify an individual, it also recognize other personal data associated with an individual such as other photos featuring the individual, blog posts, social networking profiles Internet behavior, travel patterns, etc all but only the facial feature of that person alone The technique used for face recognition are classified as traditional,3-dimensional recognition and skin texture analysis.
-
Image Acquistion
-
Image classification is to classify the pixels of an image into classes or themes. Image classification is the utmost important part of digital image analysis as its main purpose is to identify and analyze the features of an image and then portray how these features are actually represented on the ground. Image classification uses the smallest unit represented in an image i.e. a pixel for reflectance statistics. The two approaches that are used for image classification are supervised image classification and unsupervised image classification technique. Selection of image classification techniques is based on the spatial resolution of the image. The main classification of an image is on the basis of its visual content.
-
Image compression is reducing the data size of the image so that the image can be stored efficiently in lesser memory space. The image compression can be lossy as well as lossless as during compression the data can be lost and sometimes not. The compression of images during medical purposes, technical drawings etc. should be lossless as loss of small data can result in the loss of some important information whereas the image compression that is applied on natural photographs can be lossy, as the data loss during compression does not bring about any major change in the image and it also do not contains important information. The most common method that is employed for lossy compression is conversion of color space of the image into most common colors in the image. The main aim of the image compression is to obtain a quality image of the given size at a given bit rate without degradation of the image. Image compression allows less memory usage and also decreases time required to send the image over internet.
-
Image denoising is removing the noise from image and obtaining a good quality image with removed noise and preserved edges. It is one of the tasks of image processing. The prime purpose of image denoising is obtaining original image by removing or suppressing noise from the contaminated image. The noise in the image can be added manually or can be from external sources, image denoising means this added noise to be removed using some sort of filters to obtain better quality image with least or no noise. Image denoising finds its applications in the areas where good quality image is of great importance for obtaining expected results like in the areas of image registration, image restoration, visual tracking etc. The main challenge in this area is acquiring original image of high quality when noise level is high. Image denoising is the first step applied when analyzing the image data. So, it is a field of utmost importance in the area of image processing.
-
When we take an image, because of some noises or color effect ,the quality of image is degraded, so to improve the quality of image the image is enhanced ,some of image features are taken in consideration for the enhancement so that the better image is formed from the original image ,for this we use image enhancement . Image enhancement is the process that is used for adjustment of the image; by this we can enhance the feature of the image the original image and making it more suitable for display. Original image feature are enhanced like we can remove the noise, sharpen or brighten an image, making it easier to identify key feature. In image enhancement the emphasize is put on the features of the image that make the image more pleasing to the observer. Various methods of image enhancement are filtering with morphological operators ,Histogram equalization, Noise removal using a filter, Linear contrast adjustment, Median filtering etc
-
With rapid advancements in technology, it is now possible to obtain information from multi source images to produce a high quality fused image with spatial and spectral information .for this we use image fusion. Image Fusion is the process of combining two images into single image and the obtained image is more informative than the other images. The resultant image will have information of both the image, and the image that is having more information is taken. In Image Fusion is the quality of information is improved from a set of images, as images are combined to get more informative image. Some of the applications of image fusion are medical imaging, microscopic imaging, remote sensing, computer vision, and robotics.
-
Image quantization is a compression technique that is to quantize the range of values to a single quantum value in a image. This a lossy compression technique in which some amount of data is lost during transmission. In this technique the number of colors used in an image is reduced to a lower number by quantizing the color values. This technique is helpful when the image is to be displayed on a device that supports less number of colors and also for efficient compression of certain kinds of images. for equivalent compression to be applied to full image, the image is divided into blocks of dimensions 8*8. The quantization on each block is equally applied so that same amount of colors are removed from each block and blocks of resultant image formed do not appear to be quantized unequally. The quantization technique should be applied such that the blocks are not distinguishable in the resultant image. Image quantization when intentionally applied noise on it is used to prevent banding artifacts.
-
Image is restoration is the process of taking noisy image and try to clean the original image, while taking an image, sometime due to some factor like camera mis-focus, motion blur or noise the image obtained is not clear, so to restore the feature of the original, image restoration is done. Thereby increasing the quality of the image clicked. Image restoration is different from image enhancement, as in image enhancement the image feature are enhanced, so that image will become more pleasing to the observer, where as in image restoration the main focus is to obtained the clear original image which is alter due to some factors, also after the removal of noise from the image ,its resolution does not change. This is one of efficient technique of image processing to recover the image.
-
To search an image manually from the large set of data is quite time consuming, difficult and expensive, for this a system called as image retrieval is used .Image retrieval is a computer system that is used for searching and retrieving the image from large data set of digital images the methods of image retrieval consist of meta data (data about data) like captions, keywords etc so that the image us retrieved easily .To find an image, a user may provide some keywords, images, link etc. The search is made on the basis of Meta tag, color distribution in image, shape etc. image retrieval is used in many fields likes data mining ,Education, medical Imaging ,crime prevention, weather forecasting ,Remote sensing etc.
-
It is process of dividing a digital image into multiple parts called as set of pixels or super pixels. Image segmentation is done in order to make image easier to analyze and to present the image into more meaningful form than the original one. Boundaries and objects of the image are located by the Image segmentation. In this every pixel of the image is checked and the pixels that are having same label have certain common characteristic. There are various methods for image segmentation like thresholding, clustering method, Histogram based, edge detection etc. It is used in application like Machine vision, medical imaging, objection detection, Recognition Tasks, Video surveillance etc.
-
Image Steganography is the process of hiding text, image, and file into an image. Steganography is basically done to for security of transmitted data .the data that is to be transmitting secretly is hided in the image and is send to the destination. Only the sender and the intended recipient is able to detect the data sent. As the data is invisible, so it cannot be detected easily. There are many ways in which Steganography is done. The messages appear as articles, images, lists, or sometimes invisible ink is used to write between the lines. Steganography is used in modern printer. It is used by intelligence services and is allegedly used by terrorist. Steganography is different from watermarking as watermarking process in which verifies the owner is embedded into digital image whereas in Steganography only sender and recipient knows about the data that is sent.
-
Image stitching is the process of combining number of images together to produce a high resolution image. The overlaps between the images need to be exact and the exposure needs to be identical to produce high quality smooth and continuous results. Image stitching process is employed in the areas of medicine, image stabilization, video stitching, object insertion etc. the prime concern while applying the process of image stitching is that the gaps should not be visible, and the exposure between the images should be consistent. The images should be first detected on the basis of their features so that similar features can be detected automatically and the aligning of the images can be robust. The distinct features of the images are also required to be detected so that these could be matched and correspondence establishments could be rapid. The first step in image stitching is deciding the composite surface so that the other images can be stitched over it.
-
As security is the prime concern. Data that is sent from source and destination should be secure. For the security many new techniques are propose, so that data is not access by an unauthorized user. Image Watermarking is such a technique for sending data to destination without any unauthorized access i.e. by hiding data into image. Watermark is an image or a pattern that is used to check the authenticity .Image watermarking is a technique to embed data into the image. In this process the data hiding into image. It is a process in which the information which verifies the owner is embedded into the digital image or signal. These signals could be either videos or pictures or audios. Watermark in the image is used to verify the integrity or to shows the identity of its owners. No one can copy watermarked image as the watermark is copied along with image .Watermarking is used for Copyright protection, source tracking, Broadcast monitoring etc.
