The Research, Part Of The Algorithm Based On Chain Code

Chain Codes is one of the coding expressing methods for the verge of the contour of the objects. It makes full use of a series of lines with special length and direction to express the verge of the objects. On account of the limited length and the numbered directions, the jumping-off points of the verge must be expressed with absolute coordinate value, and the others' offsets can be represented by the following direction-value. The demanded byte-number of representing one direction is smaller than the one of representing one coordinate-value, and to one point, a direction-number can represent two coordinate-values, as a result, chain codes can reduce the numbers of data, which are necessary for the verge expression. The digital images are composed of griddling with fixed space between each other, so the simplest chain codes is following the verge and give the direction-value to the linked line of every two adjacent pixels.The work includes:(1) According to the chain codes, the paper gives a practical method to calculate the target area.(2) According to the chain codes, the paper gives a new method to judge the contour direction.(3) According to the chain codes, the paper gives a powerful and correctly contour filling method.(4) According to the chain codes, the paper gives a new dilation and erosion method.(5) According to the chain codes, the paper gives a practical method to detect line.Chain Codes is propitious to the calculation of characteristics of objects, such as target area, perimeter, inflexion and the pick-up line, etc. In fact, Chain code has become one of the most common tools in image and processing and pattern recognition. In this paper, we reconsidered the existing algorithm and put forward the more practical methods.