Research On Recognition And Decoding Technology Of The Coded Targets

As an important branch of machine vision,visual measurement plays a more and more important role in feature recognition and target matching.Visual measurement uptake measured target image by the image capturing device,and then the image is converted into a digital signal by the image processing system according to the pixel distribution,the brightness and color information of the target image,finally,it uses a variety of operations to process it to obtain its characteristics.Sticking the easily-identified coded targets on the measured object can significantly improve the accuracy and accuracy of feature recognition,ensure the reliability of target matching.Therefore,the identification and decoding of coded targets has always been a research hotspot in this field.The main contents of this paper are as follows:(1)Several common coded targets are introduced,and their advantages and disadvantages are analyzed.A lot of experimental verification is carried out on several coded targets.Finally,according to the needs of the subject,the circular coded target and the encircling coded target are selected as the research object,and the coding principles of the two kinds of coded targets are introduced respectively.(2)The traditional recognition and decoding algorithm of the coded targets is introduced,in view of the low efficiency of traditional coded targets recognition algorithm,a new algorithm based on improved line scan algorithm is proposed.First,the image of the coded target is processed by grayscale,denoising,grayscale stretching and binaryzation processing;secondly,the line scan algorithm is used to extract the target.Finally,the area threshold method and the shape criterion method are used to extract the circular mark points,the gray square weighted centroid method is used to locate the circular coded targets.The experimental results show that the improved line scan algorithm is used to extract the coding mark points,which obviously reduces the recognition time of the coded targets,and greatly improves the recognition efficiency and recognition accuracy.(3)For the problems of low efficiency and low recognition accuracy of decoding algorithms of the circular coded targets and the encircling coded target,the traditional decoding algorithms have been improved.For the circular coded target,this paper designs a decoding algorithm based on decoding endpoints.Decoding endpoints are obtained and decoded at the end.For the encircling coded target,this paper designs a coding and decoding method for the encircling coded target based on the template point.That is,the template points are added between the head and the tail of the original circular coded target,decode by the angle relationship between the circular marker point and the template point vector.Experimental results show that the improved recognition and decoding algorithm of two kinds of coded targets have greatly improved the recognition accuracy and recognition efficiency of the coded targets in various cases,and have strong robustness and high comprehensive performance.