A digital image is a numerical representation of (usually binary) a two-dimensional image. It could be of the vector or the raster type. Normally, the phrase “digital image” refers to raster or bitmapped images. Raster images possess a fixed amount of digital values, which are called picture elements or pixels. Every image has a fixed number of rows and columns of these pixels. The pixels are stored in the memory of computers as raster images consisting of arrays of numbers. These are stored and transmitted in compressed formats (Sprawls, 2017). Computers are able to produce descriptive information about a digital image. An example of the information that is retrieved through a digital image in the form of numbers would be the number of cells in a certain field of view of a microscope, or the diameter of a certain cell (Walter and Berns, 2013). This kind of analysis in digital imagery is called feature extraction. This process helps bring down the number of resources required to describe and analyze a large set of data (Mathworks, 2017). A digital image is usually stored by the computer in a file that consists of two main parts: the header, which contains the file format and the verification information, and the image data itself. Today, there are many methods employed in digital image processing in order to extract the information contained in the image. One such technique is image segmentation. This happens when a digital image into several segments or sets of pixels. In this technique, the image is broken down into something that is easier to analyze. This technique is normally utilized to locate boundaries such as lines and curves in images. Image segmentation assigns a label to each pixel in every image such that the pixels with the same labels have common characteristics. An application would be in the field of medicine wherein the digital image of a segmented femur, wherein the various layers are shown separately by means of color differences (Shapiro and Stockman, 2001).
The information taken from an image can be used for several purposes. Analysis can also be done for many reasons. Among the more popular reasons for analysis are office automation, automatic inspection in a manufacturing context, medical/biological analyses, remote sensing, scientific applications and in criminology. In office automation, the image analysis can aid in the processing of documents. In automatic inspection, any defects in the manufactured products can easily be identified and sent back for improvement; in the medical or biological field, analysis of images such as X-ray images using the numerical information helps physicals and other medical professionals to identify certain diseases or determine the progression of a certain disease. In remote sensing, the information taken from a digital image can provide useful information pertaining to mineralogy, forest depletion, hydrology, cartography, air travel and the like. In scientific applications, numerical information in images helps scientists make certain conclusions about outer space and other similar phenomena. In criminology, the numerical information is helpful in the areas of face recognition, fingerprint matching and forensic investigation. The numerical information from an image will also help in the area of the military for the easy identification of targets and in tasks such as missile guidance (Chanda and Mahumder, 2011).
Information retrieval is performed by first entering a query into the system of data. These queries are statements of needs of information, and several objects in the system of data may match the query initially. Thereafter, the information is sorted by relevance, by placing a score for relevancy beside each match between the query and the data. This is how the information is retrieved and prepared for analysis.
References
Chandra, B. and Mahumder, D. 2011. Digital Image Processing and Analysis. New Delhi: PHI Learning Private Ltd.
Mathworks. 2017. Feature Extraction. Retrieved from: https://www.mathworks.com/discovery/feature-extraction.html
Shapiro, L. and Stockman, G. 2001. Computer Vision. Upper Saddle, NJ: Prentice-Hall.
Sprawls. 2017. Digital Image Characteristics. Retrieved from: http://www.sprawls.org/resources/DICHAR/module.htm
Walter, R. and Berns, M. 2013. Digital Image Processing and Analysis. In Video Microscopy, S. Inoue, ed. NY: Springer Science.
