Font Size: a A A

Research And Realization Of The Target Extraction And Detection Technology Of Chip Resistors Based On Dynamic Video

Posted on:2020-09-21Degree:MasterType:Thesis
Country:ChinaCandidate:J Y ZhangFull Text:PDF
GTID:2432330602959812Subject:Control engineering
Abstract/Summary:PDF Full Text Request
The author’s research group has developed an automatic detection system for Surface defects of Surface Mounted Devices Resistor for an electronic company.In order to improve the detection efficiency of the detection system,the pre-treated equipment is designed and optimized by combining the research results of target extraction and detection technology.Finally,Surface Mounted Devices Resistor can be automatically extracted in dynamic video,and the cost of direction screening equipment can be reduced.The research and implementation of Surface Mounted Devices Resistor target extraction and detection technology based on dynamic video are carried out in this paper.The main work and conclusions are as follows:(1)I design embedded Linux video system and realize the functions of acquisition,coding and transmission.The software and hardware structure of the system is designed and the internal processing flow is realized.The embedded Linux video system’s controlling,acquisition,coding and transmission modules is built.I design the Uboot program to provide system software running environment;I build embedded Linux operating system;I build HiSilicon media processing platform and other related drivers;I realize RTSP(Realtime Streaming Protocol)network server.(2)I design and implement the function of PC video processing platform.The software program of PC is compiled to receive real-time video from RTSP network server.Extraction and Detection Computer uses target extraction and deteetion technology to process the patch resistance in video and display the results.(3)I have studied and realized the technology and process of extracting the Surface Mounted Devices Resistor.Comparing the experimental results,a more stable background elimination method is selected as the main target extraction algorithm;And a more intuitive Canny operator is selected to detect edges to get more accurate targets.Finally,the background subtraction method is decided,and the target extraction technology is realized by combining Canny edge detection.(4)I have studied and realized the target detection technology and process of the Surface Mounted Devices Resistor.Comparing the efficiency of SURF feature algorithm and SITF feature algorithm in calculating rotation angle and identifying positive and negative patch resistance,we choose SURF feature detection algorithm with little difference in accuracy but better efficiency.Finally,we decided to use SURF feature detection algorithm to achieve target detection technology.
Keywords/Search Tags:target extraction, feature detection, Opencv, HiSilicon
PDF Full Text Request
Related items