Threshold-based Image Segmentation In The Image Slips

Image segmentation by image processing to image analysis of the key steps, is also a basic computer vision technology. This is because the image segmentation, object separation, feature extraction and the parameters in the original image into a more abstract and more compact form, making more high-level analysis and understanding possible.Bamboo slips is China's unique carrier with high historical value. Because it is long buried, resulting in the text fuzzy, it was badly damaged, the character recognition difficult. Therefore, bamboo slips were segmented text, it becomes an indispensable task.The article focuses on image segmentation techniques are a general description, focusing on threshold-based segmentation method. This is mainly based on fuzzy degree of segmentation method based on genetic algorithm and maximum difference between the image on the bamboo slips for image segmentation.1. Threshold segmentation method is the most basic image processing segmentation method, which has computation, and simple to achieve, in image analysis and recognition play an important role. Image threshoiding is in accordance with the gray level, the image space is divided into scenes with reality that corresponds to some meaningful regions, each region within the gray level is uniform, while the adjacent region of gray scale is different, there remain boundary. It's divided by starting from the gray level, select one or more threshold values to achieve.2. From bamboo slips image's characteristics, select the degree based on fuzzy threshold algorithm, consider this method of artificial selection of bandwidth limitations, the introduction of fuzzy adaptive threshold algorithm. It can be automatically selected according to the histogram peak distribution of the size of window width, greatly improved the efficiency of image segmentation bamboo slips; Moreover, taking into account the bandwidth selected dependent histogram peak distribution histogram peaks that do not have the characteristics of image segmentation bamboo slips will result in failure to adopt a histogram transform this new approach to solve this problem, so most of the bamboo slips image to achieve the desired segmentation.3. In Otsu theory, image segmentation in the bamboo slips into genetic algorithm. Focuses on the genetic algorithm based on the maximum difference between two classes of image segmentation As the Otsu algorithm not only consider the image intensity data, but also to consider the field of spatial information, can guarantee the accuracy of segmentation, while the genetic algorithms can greatly improve the speed threshold optimization. Therefore, the combination of the two methods split bamboo slips image, not only to improve the searching speed, but also to ensure the accuracy of segmentation, to segment bamboo slips well.