Research On The Identification Method Of Ground Dew Condensation Based On Hough Transform

There are many weather phenomena in nature,including but not limited to rain,snow,dew,frost,hail,etc.With the development of science and technology,for a variety of needs,the demand for the observation of these natural weather phenomena is also increasing.The ultimate goal of weather phenomena observation is generally all weather,all weather,automatic and accurate observation,in which automatic and accurate observation becomes the focus of research.This paper uses machine vision to automatically identify condensation air phenomenon.The possibility of dew shape characteristics in dew recognition is studied,an observation system for collecting dew data is designed,the advantages and disadvantages of different preprocessing,edge detection,and recognition algorithms for observation images are analyzed,an identification processing algorithm based on shape features is proposed,and the collection method of dew coverage information is studied on the basis of shape feature judgment.The specific contents of this paper are as follows:(1)Dew observation system design: In order to solve the problem of dew data collection,a method of using artificial condensation contact surface instead of natural condensation contact surface is proposed,and then the data is collected by the camera and sent to the client for storage and processing,which enables the automatic acquisition of dew condensation image on the ground.(2)Preprocessing of ground dew observation image: In order to make the dew image acquired can be subsequently extracted accurately,the effects of different processing algorithms on the original dew image are discussed,the results of different algorithms are compared,and the pre-processing of dew image is achieved.(3)Feature detection based on the shape features of the observation image: The shape features of dew in the observation image are studied,and the Hough transformation is used to detect the circular shape features of dew in the observation image.Canny edge detection is used to extract the edges of the observation image before subsequent feature detection.It can suppress noise while preserving image edge information and reduce unnecessary details of the observed image.The detection accuracy is 91.8%.(4)Dew coverage detection: The pixel area of the observation empty carrier is proposed,and the area and sum of the observed circles are solved.Then the ratio is used as dew coverage information reflecting the condensation amount of dew.The error between the ratio and the manual comparison is only 11.8%,thus diversifying the collection of dew data.