In image processing, improving the quality of the digital image by some manipulations is known as image enhancement. Improving the image involves manipulation in contrast, changes in the brightness of the image. The digital image undergoes various modifications including contrast stretching and noise elimination such as Poisson noise, speckle noise, salt and pepper noise, Gaussian noise. Filters are used to eliminate noise in order to improve image quality. Image processing is a method in which certain operations are applied to the digital image in order to extract some useful information. It implies the number of phases in which the improvement of the image is one of them. It refers to sharpening the characteristics of the image as manipulating the contrast for clear visibility, improving the limits. The enhancement of the image is done without the knowledge of the degradation of the image. If degradation is known, it is called image restoration. Many different methods are used to improve the quality, generally elemental and heuristic method. The method to be used depends on the application, ie where the image is used because there is no single quality measure. One method used for one problem may not be suitable for another. Therefore, the techniques are problem-oriented. Two widely-classified types of methods are spatial methods and frequency methods. Spatial domain methods are applied directly to the pixels in the image. The techniques are gray level cutoff, histogram equalization, point processing operation, negative of an image, logarithmic transformation, power law transformation. Frequency domain methods are based on modifying the spectral transformation of an image. The techniques are Fourier transform, filtered, homomorphic filtered.
Limitations of image improvement
Improving the image as an aid to the visually impaired can improve the visibility of television programs and provide portable visual aid. This article describes current techniques for image enhancement and its underlying models. The limitations of the various techniques and possible methods of application are highlighted. The initial work in this area was based on a linear model. The finite dynamic range available on the video display and image contamination enhanced by high frequency spatial noise limited the utility of the model. I propose a method to address some limitations of the original model that considers the nonlinear response of the visual system and requires the increase of sub-threshold spatial information only. This modification can increase the available dynamic range by decreasing the range previously used by linear models to improve the visible details. However, for the modified technique to be more effective, the improvement must be adjusted continuously, based on the visual loss of the patient and the spatial frequency content of the visualized images. The implications of these limitations for potential implementation on TV are discussed. Implementing a visual enhancement aid on a binocular, full-field, head-mounted virtual vision device can cause substantial difficulties. Adaptation of the patient may be difficult due to head movement and interaction of the vestibular system response with the head-mounted display. An alternative bioptic design is proposed in which the screen is placed above or below the line of sight to be examined intermittently, possibly in a frozen frame mode. This implementation is likely to be less expensive, allowing more users to access the device.