Extraction of the small details on the noisy images and their sharpening: implementation on the CNN
Igor N. Aizenberg · 2002
Applications of the algorithms of impulse noise filtering, edge detection, high and medium frequencies amplification to improve the quality of gray-scale images, especially of satellite images, are considered. All the algorithms presented are implemented on the cellular neural network (the classical CNN type or based on universal binary neurons). The strategy of processing small-detailed images using the noise filtering without smoothing, edge detection and extraction of the small or other important details on the complex image background, has been carried out.