Survey about Shape Detection

Abstract

The simplifying original shoreline of vector data into direction-line is the source of picture of baselines of the shape detection. Applying standard algorithms of vector data confining, such as the Douglas-Peucker algorithm, to analyze the real shoreline within direction-line, disregards the specific elements of a guarantee of general marine interests .To solve this problem, this research suggested an improved Douglas-Peucker algorithm. First, determine the curved vertices of the original shoreline as a choice set of points; later filter division points from curved vertices based on corner and length; ultimately, separate the data into different parts in order to complete the Douglas-Peucker algorithm, and compare the threshold of each part to efficiently select specialty points and protect shape detection. This research uses Python language to carry out the algorithm design, and openCV2. The results show that, compared with the Douglas-Peucker algorithm, the improved Douglas-Peucker algorithm improves how detect the contours, and shape detection, and the accuracy of curve simplification also fixed.