Monitoring Method And System For Uniform Powder Mixing Based On Image Technology

Barrel finishing process is a basic technology in the field of manufacturing,which is used to improve the surface quality and performance of parts.Barrel finishing process has advantages in terms of environmental friendly,parts performance,adaptability and economic.The abrasive media is the abrasive tool in the barrel finishing process,which has an important influence on the effect of parts processing.The sintered abrasive media is the most widely used abrasive media.The mixing uniformity of powder materials is the key factor affecting the quality during the preparation of sintered abrasive media.If the powder materials are not mixed uniformly,not only the quality and service life of the abrasive media will be reduced,the abrasive media is prone to delamination,deformation,cracking,sticking,etc,which will cause abrasions or scratches on the surface of the processed parts.At present,the detection of the mixing uniformity of powder materials during the preparation of sintered abrasive media mainly relies on the experience,with large error,high rejection rate,poor processing effect and short service life.Regarding the problems above,this paper proposes to judge the mixing uniformity of powder raw materials by the degree of similarity between the images of the raw powder materials.And using C # language to design a relevant platform interface of the powder mixing uniformity monitoring system based on image technology.The main contents of this paper are:(1)The powder sampling system is designed to collect powder samples and images by sampling device and digital microscope.The appropriate parameters are selected by analyzing the factor of image resolution,magnification,and light intensity on the image effect.The collected powder images are preprocessed with graying,downsampling and sharpening.(2)For the pre-processed powder image,the composite color feature parameters is used as powder image features.Firstly,the characteristic parameters of the powder images are counted by color histogram,and the characteristic parameters of RGB and HSV space in the color features are combined as composite feature parameters.Then the principal component analysis is used to reduce the dimension of image features to remove the cross redundancy information.Finally,the characteristic sample matrix is constructed after data normalization.The simulation results show that the dimension reduction composite feature parameters selected in this paper can effectively represent the images of the powder images.(3)The variable weight euclidean distance is used to accurately and effectively detect the similarity of powder images,and to judge the uniformity of powder mixing.Firstly,the sample density center is calculated as the central feature.Then the entropy method is used to assign weights to different characteristic parameters and to determine the weights of each principal component.Secondly,the similarity degree of varying weight from each powder images to the central feature of each group of images is calculated,and the minimum similarity value is taken as the similarity of the group of images.Finally,the appropriate threshold value is selected to judge the mixing uniformity of powder by combining with the expert knowledge in the field of barrel finishing,the characteristics of different mixing powder and the existing methods for calculating the mixing uniformity in various fields.The results show that the method can accurately and effectively determine the powder mixing uniformity.(4)The interactive interface of the powder mixing uniformity monitoring system based on image technology is designed by the C#.And the process parameters of the powder sample image acquisition,image pre-processing,feature extraction,and similarity detection are displayed in real time.