Fast selecting threshold algorithm based on one-dimensional entropy

According to the characteristic that uniform probability distribution of gray levels maximizes the Shannon entropy, we define a new objective function of simple form and definite meaning to select threshold. The threshold which is selected by the new objective function is the same as the one that one-dimensional entropy-thresholding. The character will be proved theoretically and validated experimentally in this paper. Although the two thresholds obtained by using the two methods above are same, but the computational time are very different. Because, the new objective function only uses subtraction operation to select threshold, and the one-dimensional entropy-thresholding uses logarithm and product operations, so the method proposed in this paper takes less computational time than one-dimensional entropy thresholding, The experimental results show further the computational time is decreased by 15.53% at least. In a word, the method proposed in this paper can be considered as a fast algorithm of one-dimensional entropy thresholding, also has the same segmentation effect as the one-dimensional entropy thresholding.