In a typical digital image processing system, the source of the image is usually visible light reflected from or transmitted through various objects in a scene. Optics gathers and focuses this light into a sensor that puts an electronic signal to the received light. Images can also be formed using other sources of radiation such as infrared or ultraviolet light, X-rays, radar, or sonar. Images can be synthesized from spatial data by other means, including scanning and computer-aided tomography.

The sensor signal is "digitized"--converted to an array of numerical values, each value representing the light intensity of a small area of the scene. The digitized values are called picture elements, or "pixels," and are stored in computer memory as a digital image. The limited range and number of pixels means that the digital image is only an approximate of the light inte

Often the recognition can be made more reliable by using "feature detectors" or "matched filters" to amplify or find specific image features that contain unique or important information about the objects. The resulting features, measurements, or images are examined for patterns that match the various objects. This examination might use pattern analysis methods to recognize reliably the objects in the image.

Image processing is used to edit images, perhaps for use in a magazine. Image editing uses many of the methods from image enhancement and restoration, such as removing image blur (or adding blur) and changing the location of pixels. For example, an element in an image might be "cut" out, reduced in size, and inserted ("pasted") in another image. The edges of the inserted image differ in intensity and might be noticed. To remove this unwanted edge, the inserted image is smoothly blended (blurred) into the background by averaging pixel-intensity values across the edges.

Image-recognition algorithms attempt automatically to find and identify parts or objects within an image. Typical recognition tasks are not this simple, and often require finding and recognizing objects in cluttered and degraded images. One recognition method compares images of the objects with every area of a sample image. If a template matches some area of the sample image, the image might contain the corresponding object. Unfortunately, the match is usually imperfect due to image noise, object variation, object rotation, changes in lighting, and other factors and so statistical methods are used to decide if the match is valid.

Image compression reduces the amount of information required to store

Some common words found in the essay are:
Scans Images, RECOGNITION Image, RESTORATION Image, EDITING Image, COMPRESSION Image, X-rays CAT, John Kennedy's, digital image, image processing, image enhancement, image image, image compression, values image, image editing, image analysis, light intensity scene, various objects, human vision, image enhancement restoration,
Approximate Word count = 1133
Approximate Pages = 5 (250 words per page double spaced)