Constructing And Implementing For Non-touch Measurement System Of Cattle Body

With the rapid development of electronics, optics, computers, machinery and otherindustries, photoelectric detection technology that is characterized by non-touch,high-precision, high-speed has become the main detection technology development. Becauseof inconvenience of traditional measurement techniques, the non-touch measurement methodhas been loved by many researchers.Non-touch measurement do not contact with the object tobe measured directly, as opposed to manual measurement, which has convenient,high-efficiency, high-precision features, and avoiding the impact of man-made factors.Currently, non-touch measurement technique is widely used in astronomy and humanbody research, but seldom used in agriculture. To solve this problem, this study mainly usesnon-touch measurement methods to measure body size of ox, in order to understand thevarious parts of the cow body growth and development, and nutritional status.In this paper,for the inconvenience of current traditional manual measurement methods of cow body, usingcomputer image processing, establishing a non-touch cattle body measurement system.According to the characteristics of non-touch measurement,using two cameras shooting thesame cattle from different angles,and using binocular stereo vision technology for imageprocessing, building and achieving a non-touch measurement system of cattle body.The chiefwork is as follows:(1) Obtaining and analyzing the image of cattle bodyThe principle of binocular stereo vision, using binocular camera shoot the same cattlebody,getting the two images of the same body, and numbered and saved;According to thedifferent No. of the cattle body, extracting images for measuring.Matching the correspondingpoint of the same cattle body in the two images,entering the measurements.(2) Measuring and calculating the cow body measurementsMeasuring the cattle body in the image after processing,obtained the actual size of cattlebody by calculating and saving the data.Firstly,using the Euclidean distance calculating the length of the two-dimensional in pixels between two points, obtaining the two-dimensional data by proportional relationship.Using the known Bezier curve data reverse obtaining the control point to obtain thethree-dimensional,and calculating the distance between the control point to get the curvelength.(3) The design of software is completed using Qt cross-platform C++graphical userinterface application development framework, combined with OpenCV （Open SourceComputer Vision Library）open source computer vision library. Using MySQL database,according to the picture number to get a picture displayed on the screen, and saving themeasured data.