Target Tao Experimental Image Automatic Interpretation |
Posted on:2006-08-22 | Degree:Master | Type:Thesis |
Country:China | Candidate:Y Xu | Full Text:PDF |
GTID:2208360155958978 | Subject:Artificial intelligence and pattern recognition |
Abstract/Summary: | PDF Full Text Request |
Feasible method of digitalizing shadow photograph of range and using advanced image recognition techniquc to recognize mass center position and posture of projectile in ballistic range is studied and the processing procedure is proposed. First the projectile's region and two reference triangle regions in the shadow photograph are evaluated automatically by image segmentation technique. Then three regions are transferred into binary images respectively by an improved valve binary algorithm. The pure binary images are obtained by filtering noises.The optimal objects are acquired by dilation transform and erosion transform from mathematical morphology. Finally,according to the transcendental knowledge of the projectile, the accurate coordinate and posture angle of projectile are calculated. Using the automated reading system to process a group of spark shadow photographs of one kind of shell. The result shows that using image recognition technique can improve the precision and efficiency of photographs reading greatly and has great application value in ballistic range experiment. |
Keywords/Search Tags: | Digitalize, Image recognition, Shadow photograph of range |
PDF Full Text Request |
Related items |