Method for detecting oblique triangle in digital image

Abstract

The invention relates to a method for detecting oblique triangle in a digital image. The method comprises the following steps of: acquiring the image and inputting into a computer; computing the gradient of each pixel point; computing an edge graph of the image; computing the directional line of each edge point by using the gradient of each edge point on the edge graph; computing the characteristic length and the characteristic energy of each pixel point in the image to acquire a characteristic length allocation graph and a characteristic energy allocation graph of the image; detecting local maximum points on the characteristic energy allocation graph under the limitation of a threshold value; determining a set of edge points which form a triangle according to each local maximum point and the characteristic length corresponding to the local maximum point; grouping edge points for the triangular edge point set according to gradient directions; verifying and removing unreasonable triangular edge point sets by using grouping numbers; fitting a straight line for each group of edge points of the grouped triangular edge point sets; acquiring the vertex position of the triangle by intersecting straight lines; and finally outputting the vertex position information of the triangle. By the method provided by the invention, the triangle in the image can be exactly detected; and the method is easy to realize, does not need information relative to the given image and has higher computational efficiency compared with the conventional method.

Description

technical field [0001] The invention relates to the field of automatic detection of image features in computer vision, in particular to a detection method for arbitrary triangles in digital images. Background technique [0002] Shape detection and recognition play a very important role in the fields of automatic detection, object positioning, image analysis, and computer-aided design. At present, there are many methods for identifying closed geometric figures composed of straight line segments such as polygons. Generalized Hough Transform (GHT) [1] Using the geometric properties of polygons, the detection problem of variable space graphics is transformed into the clustering problem of parameter space, and the direct detection of polygons is realized. It is characterized by simplicity and directness, but it has a large amount of calculation and low calculation efficiency. Lara, etc. [2] A parallel algorithm is proposed to recognize polygons in the image. Under the conditi...

AI Technical Summary

Problems solved by technology

Barnes et al. [3] The proposed method first obtains the edge of the image, and then uses the posterior probability to define the probability density function of the regular polygon according to the geometric characteristics of the regular polygon, and then realizes the detection of the regular polygon in the road sign by calculating the number of sides and the direction deviation of the regular polygon. The calculation of the method is more complicated, the calculation amount is large, and it is only for a specific polygon

Images