Introduction to Image Processing
Image processing is a method to perform some operations on an image, in order to get an enhanced image or to extract some useful information from it. It is a type of signal processing in which input is an image and output may be image or characteristics/features associated with that image. Nowadays, image processing is among rapidly growing technologies. It forms core research area within engineering and computer science disciplines too.
Image processing basically includes processing of images using mathematical operations by using any form of signal processing for which the input is an image, such as photographs or frames of video; the output of image processing may be either an image or a set of characteristics or parameters related to the image. Most common operations performed during image processing include:
Image acquisition – Involves inputting the image using scanning, shooting using digital camera etc.
Image enhancement – It aims to improve the quality of image or to highlight some features for further processing. E.g. histogram equalization, filtering etc.
Image segmentation – It partitions/divides image into multiple segments sets of pixels.
Image representation and description – It allocates appropriate properties to represent image content.
Image compression – It reduces the storage requirements and transmission bandwidth needed for an image.
Image recognition and interpretation – Identify and interpret the objects, structures, or meaning within the image.
Quantitative analysis – Measurement, counts or other quantitative information directly from image itself.
In nutshell, main applications include medical imaging such as X-rays, nuclear medicine, ultrasound etc., remote sensing, non-destructive testing, military/defence etc. Let us now discuss some key aspects of image processing in detail.
Image Processing Steps
The major steps involved in digital image processing are:
Image Acquisition: Digital image is obtained with help of devices like scanner, camera, satellite, etc and is stored in the memory of computer systems.
Preprocessing: This step is performed to eliminate noise if present in the image and to improve the quality of image. Operations like filtering, cropping, enhancement are done.
Segmentation: It divides the image into groups of pixels based on color or texture for easy identification of objects and background.
Representation and Features: The segments are represented in some data structure like arrays for further processing. Useful features representing each object in the image are extracted.
Description: Attributes along with topological and geometrical information is extracted for high level description of image contents.
Recognition: It identifies objects in the input image by use of pattern recognition techniques.
Interpretation and Analysis: The input image is interpreted and analyzed based on the semantic meaning of the objects recognized for applications like medical diagnosis.
Post Processing: At the end, information extracted is displayed on the screen. Compression is done for storage or further processing.
Major Techniques in Image Processing
Some of the major techniques used in image processing are as follows:
Image Enhancement Techniques: It aims to process an image so that result is more suitable than original image for a particular application. Examples are histogram equalization, filtering etc.
Image Restoration Techniques: It tries to build degraded/blurred image for example due to noise, imprecisions in sensor etc. Example is smoothing filter.
Image Compression Techniques: It aims to reduce irrelevance and redundancy of image to store or transmit data efficiently. Example is JPEG, MPEG etc.
Image Segmentation Techniques: It partitions images into regions or objects. Example is thresholding, region growing etc.
Morphological Techniques: It is based on shape of image object. Example is erosion, dilation etc.
Registration Techniques: This registers two images or aligns multiple images. Example is template matching.
Object Representation and Description Techniques: It represents objects identified from segmentation process. Example is chain code, quad tree etc.
Motion Estimation Techniques: It determines motion of objects between frames for video signals. Important examples are block matching, optical flow.
Some Key Areas of Image Processing Research
Following are some key areas where active research is going in field of image processing:
Medical Image Processing: Used for applications in X-rays, CT, MRI scans etc. Research areas are image guided surgery, microscopy images analysis.
Face Recognition: Used in biometric authentication systems. Active research in illumination invariant face recognition.
Remote Sensing: Used for satellite images, drones etc. Research includes change detection in land use, atmospheric and environmental research.
Industrial Imaging: Used in applications like inspection, process control. Research areas are defect detection, automatic visual inspection.
Forensic Analysis: Used for fingerprint, document analysis etc. Also advances in video/image forensic analysis and compression artifacts analysis.
Document Image Processing: Used for applications like digitization of documents. Research areas include document image binarization, analysis and recognition of scanned documents.
Deep Learning based Applications: Advances in deep learning revolutionized areas like object detection, semantic segmentation, GANs etc.
Multispectral and Hyperspectral Imaging: Beyond normal RGB images, analyzing images captured at multiple wavelengths. Research in advanced applications like mining, defense etc.
With growth in areas like AI, big data, IoT; image processing is witnessing advancements at fast pace. It is increasingly getting applied to domains like autonomous vehicles, surveillance, smart cities too. Active research continuously pushes boundaries of this technology.
Conclusion
In this article, we discussed fundamentals of digital image processing including common terminology, important steps and techniques. We also covered some major application areas and highlighted key ongoing research domains. Overall, image processing plays integral role in computer vision systems and forms backbone of advanced technologies like self-driving cars, drones etc. With continued research and innovation, its scope and potential applications will keep expanding in future.
Andrew Stroud