CHAPTER – 2 Image Processing Basics










        Digital image processing is an important application of 2-D digital signal processing theories. It has many practical applications. One of the earliest applications was processing images from the Ranger 7 mission at the Jet Propulsion Laboratory in the early 1960s. The imaging system mounted on the spacecraft had a number of constraints imposed on it, such as size and weight, and images received had such degradations as blurring, geometric distortions, and background noise. These images were successfully processed by digital computers. and since then images from space missions have been routinely processed by digital com­puters. The striking pictures of the moon and the planet Mars we see in magazines have all been processed by digital computers. Here are some other important applications of digital image processing as follows;


1.The image processing application which probably has had the greatest impact on our lives is in the field of medicine. Computed tomography is used routinely in many clinical situations, for example—detecting and identifying brain tumors. Other medical applications of digital image processing include enhancement of x-ray images and identification of blood vessel boundaries from angiograms.


2.Another application, much closer to home for the average person, is the improvement of television images. The image that we view on a television mon­itor has flickering, limited resolution, a ghost image, background noise, and motion crawling due to line interlace. Digital televisions are not far from realization, and digital image processing will have a major impact on improving the image quality of existing television systems and on developing such new television systems as high definition television.


3.One major problem of such video communications as video conferencing and video telephone has been the enormous bandwidth required. A straightforward coding of broadcast quality video requires on the order of 100 million bits per second. By sacrificing quality and using digital image coding schemes, systems that transmit intelligible images at bit rates lower than 100 thousand bits per second have become commercially available.


4.Robots are expected to play an increasingly important role in industries and homes. They will perform jobs that are very tedious or dangerous and jobs that require speed and accuracy beyond human ability. As robots become more so­phisticated, computer vision will play an increasingly important role. Robots will be asked not only to detect and identify industrial parts. but also to “understand” what they “see” and take appropriate actions. Digital image processing will have a major impact on computer vision.


5.In addition to these well-established application areas of digital image pro­cessing, there are a number of less obvious ones. Law enforcement agents often take pictures in uncooperating environments, and the resulting images are often degraded. For example, snapshots of moving cars’ license plates are often blurred; reducing blurring is essential in identifying the car. Another potential application is the study of whale migration. When people study the migratory behavior of lions, tigers, and other land animals, they capture animals and tag them on a convenient tail or ear. When the animals are recaptured at another location, the tags serve as evidence of migratory behavior. Whales, however, are quite difficult to capture and tag. Fortunately, whales like to show their tails. which have features that can be used to distinguish them. To identify a whale, a snapshot of its tail, taken from shipboard, is compared with a reference collection of photographs of thousands of different whales’ tails. Successive sightings and identifications of an individual whale allow its migration to be tracked. Comparing photographs, though, is extremely tedious, and digital image processing may prove useful in the task.


        The potential applications of digital image processing are limitless, in ad­dition to the applications discussed above; they include home electronics, astronomy, biology, physics, agriculture, geography, defense, anthropology, and many other fields. Vision and hearing are the two most important means by which humans perceive the outside world, so it is not surprising that digital image processing has potential applications not only in science but also in any human endeavor.

        Digital image processing can be classified broadly into four areas depending on the nature of the task. These are;


1.Image enhancement :images either are processed for human viewers, as in television, or are preprocessed to aid machine performance as in object identification by machine.


2.Restoration : an image has been degraded in some manner such as blurring, and the objective is to reduce or eliminate the effect of degradation. Image restoration is closely related to image enhancement. When an image is degraded reducing the image degradation often results in enhancement. There are, however some important differences between restoration and enhancement. In image res­toration, an ideal image has been degraded and the objective is to make the processed image resemble the original as much as possible. In image enhancement the objective is to make the processed image look better in some sense than the unprocessed image. To illustrate this difference, note that an original underrated image cannot be further restored, but can be enhanced by increasing sharpness.


3.Coding : one objective is to represent an image with as few bits as possible preserving a certain level of image quality and intelligibility acceptable for a given application such as video confer­encing. Image coding is related to image enhancement and restoration. If we can enhance the visual appearance of the reconstructed image, or if we can reduce degradation from such sources as quantization noise from an image coding algo­rithm, then we can reduce the number of bits required to represent an image at a given level of image quality and intelligibility


4.Undersanding : the objective is to symbolically represent the contents of an image. Applications of image understanding include computer vision, ro­botics, and target identification. Image understanding differs from the other three areas in one major respect. In image enhancement, restoration, and coding both the input and the output are images, and signal processing has been the backbone of many successful systems in these areas. In image understanding the input is an image, but the output is typically some symbolic representation of the contents of the input image. Successful development of systems in this area involves both signal processing and artificial intelligence concepts. In a typical image under­standing system, signal processing is used for such lower-level processing tasks as reduction of degradation and extraction of edges or other image features, and artificial intelligence is used for such higher-level processing tasks as symbol ma­nipulation and knowledge base management.