Automatic Determination of Yarn Hairiness Length Based on Image Processing and Analysis Algorithm
A new algorithm is proposed to determine the actual length and the number of the protruding fibres of yarn based on a combination of image acquisition technology. First, a yarn hairiness image is obtained by the series of image processing procedures that include grayscale transformation, skew correc...
Saved in:
| Published in | 东华大学学报：英文版 Vol. 33; no. 4; pp. 587 - 591 |
|---|---|
| Main Author | |
| Format | Journal Article |
| Language | English |
| Published |
31.08.2016
|
| Online Access | Get full text |
Cover
Loading…
| Summary: | A new algorithm is proposed to determine the actual length and the number of the protruding fibres of yarn based on a combination of image acquisition technology. First, a yarn hairiness image is obtained by the series of image processing procedures that include grayscale transformation, skew correction, yarn binary image acquisition and yarn core binary image obtaining. Then. the hairiness is realized in single pixel width by the usage of thinning algorithm. Finally, a baseline of yarn core margin is obtained, and pixels that match 8- neighbor template correctly are found by row scanning in a certain area. From this, these pixels are judged and the real crossover points of yarn core margin and hairiness, i. e. , the starting points of hairiness, are gained. The real length of the protruding fibres is calculated by tracking hairiness from the starting point constantly. |
|---|---|
| Bibliography: | A new algorithm is proposed to determine the actual length and the number of the protruding fibres of yarn based on a combination of image acquisition technology. First, a yarn hairiness image is obtained by the series of image processing procedures that include grayscale transformation, skew correction, yarn binary image acquisition and yarn core binary image obtaining. Then. the hairiness is realized in single pixel width by the usage of thinning algorithm. Finally, a baseline of yarn core margin is obtained, and pixels that match 8- neighbor template correctly are found by row scanning in a certain area. From this, these pixels are judged and the real crossover points of yarn core margin and hairiness, i. e. , the starting points of hairiness, are gained. The real length of the protruding fibres is calculated by tracking hairiness from the starting point constantly. 31-1920/N protruding fibres of yarn; image processing procedures; hairiness thinning ; template matching ; tracking hairiness |
| ISSN: | 1672-5220 |