Text Location In The Nature Image

With the development of the people's life, the handle telephone and digital camera can get the nature image with word and change the image word into the edited data word. So the recognition of word in nature image has become the impending demand. To the complex nature image, the text location is the precondition of word recognition. Based on the scientific research project, we emphases on word's color information and shape information, and bring forward several useful arithmetic of text location: (1) The arithmetic of text location on RGB threshold clustering. In this arithmetic, we locally binarizate the RGB weight of color image, form several children images by color clustering, brush the image and analysis the interconnected domain in the children images. At last we synthesize all children images to get the location of text. (2) HSV clustering arithmetic. Based on the vision theory on color, we use the HSV model to divide the nature image into red,orange,yellow,green,cyan,blue,purple and black,white,gray children images, and then dong the latter work. In the improved dividing method, the image is divided into red,yellow,green,blue and gray images. (3) The arithmetic on Mixed model of HSV and Sober edge detection. It considers not only the color information, but also the edge and shape of word. And it combines these to mixed model to get the best result.The result of experiment shows that the arithmetic presented in this paper is novel and it can well and truly locate the text, and more it can fix the rough frame of word. This arithmetic is valuable in theory and application.In addition, some applicable arithmetic is introduced latter in this paper. The detection and surveillance to the bridge's crack is deeply researched, at the same time, the composing and the recognition process of analog measuring instruments system by image processing is introduced in detail.