Image Processing is one of the major applications of ANN’S which leads to the result of better and clear image which is also eye soothing for the users.
During the last decade, a significant progress in both the theoretical aspects and the applications of neural networks on the image analysis, and processing, has been made. In this paper, basic neural network algorithms as applied to the imaging process as well their applications in different areas of technology, are presented, discussed, and analysed. Novel ideas towards the optimization of the design parameters of digital imaging sensors utilizing neural networks are presented.
Digital Imaging is a process which aims to recognize objects of interest in an image by utilizing electronic sensors and advanced computing techniques with the aim to improve image quality parameters. It contains intrinsic difficulties due to the fact that image formation is basically a many-to-one-mapping, i.e., characterization of 3-d objects can be deduced from either a single image or multiple images. Several problems are associated with low-contrast images, blurred images, noisy images, image conversion to digital form, transmission, handling, manipulation, and storage of large-volume images, leading to the development of efficient image processing and recognition algorithms. Digital imaging or computer vision involves image processing and pattern recognition techniques. Image processing techniques deal with image enhancement, manipulation, and analysis of images.
These methods arise from two principal application areas:
- a) Improvement of image content for human interpretation and processing, and
- b) Processing of scene data for machine perception.
Some of their image processing methods include:
- Digitization and compression
- Enhancement, restoration, and reconstruction, and
- Matching, description, and recognition.
Digital Imaging Systems:
Digital systems with increased contrast sensitivity capabilities and large dynamic range, are highly desirable. By defining contrast as the perceptible difference between the object of interest and background, the contrast sensitivity of an imaging system is the measure of its ability to provide the perceptible difference. It can be an operator dependent or independent parameter. In this study, the observer independent contrast sensitivity was measured. Also, it is very important that a detector system is capable to record a wide range of signals coming off the object. The dynamic range provides quantitative measure of detector’s system ability to image objects with widely varying attenuating structures. It is defined as the ratio of the maximum signal to the minimum observable image signal.
‘DR=S max /∆Smin’
Where DR is the dynamic range, Smax is the maximum signal from the detector before saturation or non-linearity occurs and ∆Smin is the minimum detectable signal above the noise threshold.
Digital Imaging Applications:
Several digital imaging techniques have been developed for a large gamma of applications, such as aerospace, surveillance, sub terrestrial, marine imaging, and medical imaging applications. Applications range from imaging systems in the visible and infrared through x-rays, MRI, and ultrasound, sonar, and radar applications.