Image Of The Projectile Trajectory Analysis And Research

The target recognition has been applied in many fields, which include intelligent transportation, industrial production, community safety and military, as an important topic in computer vision systems. The target recognition algorithm has important theoretical value and practical significance.Through studying modern image processing techniques, which include image filtering, image enhancement, image segmentation under actual situation, the task of this thesis was improved in research of image pre-processing of infrared images projectile trajectory. The threshold segmentation and edge detection in image segmentation based on technology of traditional threshold segmentation were researched by designing segmentation algorithm, which can avoid shortcomings of global threshold inapplicable was caused by image bright change based on multi-window gray level threshold. The result show that this algorithm has achieved. After threshold segmentation, in order to curb left discrete noise in the binary image after threshold segmentation, analyzing the performance and characteristics of the discrete noise in the binary image and designing iterative de-noise algorithm, which based on a binary image of the neighborhood. The experiment proved that this algorithm can effectively suppress noise left over from the target image of the subject in the image segmentation point.According to linear characteristics of the projectile trajectory, the Hough Transform was researched in infrared target image of projectile trajectory detection. The characteristics segmentation method, based on the character of the vertical density projection, was designed by studying the digital character recognition technology in infrared image of the character information recognition. The results show that characteristics segmentation method has good effect of segmentation. The character recognition method of geometric combinations was designed by prior knowledge of topics character in the character recognition stage. This method has a lot of advantages that include high processing speed and recognition rate. Finally, the related study, which include the composition of the projectile trajectory image detection system, development environment and technology, as well as some of the features, were introduced in the thesis.