Research And Application Of Feather Size Measure

The application of machine Vision is widely, many sectors of the community have used the machine vision technology. A badminton have sixteen feathers, If these feather shapes vary greatly, not only affects the appearance and quality of badminton, but also there will be flying unstable conditions during the flight. In traditional feather detect process, generally, in traditional feather detecting system, feather size are detecting by fixed every pixel’s size and captured the feather’s pixel through a camera. This method ignore the bending characteristic of feather, the bending feather could make its image in camera become bigger or similar. If use the current pixel proportion size, it would bring some error in detecting process. The method in this paper use double camera to collect image, and take iterative calculate to correct the parameter of feather. This method solved the problem in traditional feather detecting system to a certain extent. The innovation of this paper on detect feather is summarized as follows:1. This paper used the Image Cut method to extract feather pole method, could analysis the shape of feather more precisely. Feather contain feather leaf and feather pole two piece, this paper used OSTU and Image Cut two method in different occasion to separate feather image. Extract the feather pole and analysis it independently, system could found the cusp of feather pole instead of feather cusp. This method could expand the scope of that can be measured, improve the adaptability of the system.2. This paper use iterative calculate method improve the old method of detecting curve value and camber value, it make the result more precisely than previous method. Feather top section have some curvature and camber, this will cause the feather cusp deviate the flat which has been calibrated, therefore, this will cause mistake to the actual measurement. This paper use iterative calculate method compensate the object distance, solve the problem of feather cusp deviate the calibrated flat.3. This paper designed a platform of detecting father thickness, in this platform, system could used one camera work, and could prevent mistake of thickness because of the different thickness of feather.The image processing system in this paper was programmed based on Matlab and VC, realized the perfect detect processing, has some theory meaning and good use value.