Digital Technique by Box Counting Method for Predicting Subjective Estimation of Good Scenery

This paper presents a digital technique for simulating image processing algorithms performed in a human perception on estimation of good scenery. The technique evaluates the landscape quality in the way of digital box counting method. In order to define good scenery, 15 examples of the paintings are selected from world-widely well known painters of the impressionists' school all acknowledged as ideal samples of good landscape scenery. When fractal dimension analysis is applied to the samples, we found that in every case the value of fractal dimension is always within the range of 1.400 to 1.700 in the case of 2D samples. We conclude that the digital image process is very useful for simulating human perception's work and the special range of values in fractal dimension may reveal equivalently human perceptions reaction for comportable feeling given by good scenery.