We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
基于条纹投影的苹果果梗/花萼识别方法.
- Authors
蔡家旭; 王飞; 奚冬冬; 刘路; 王玉伟
- Abstract
To avoid the interference of apple stem/calyx area on its surface defect detection, an apple stem/calyx recognition method is proposed in the present assay. Firstly, by taking use of phase-shifting fringe projection technology, the deflection phases of the reference plane under multiple angles are simulated with mathematical transformation. Secondly, the phase-height images under multiple angles are obtained by subtracting the deflection phases from the absolute phase of the apple. Based on the concavity characteristic of the stem/calyx, the phase-height images are filled with morphological grayscale algorithm to obtain the phase filled images. Finally, the phase difference images are computed by subtracting the phase-height images from the phase filled images, then the stem/calyx area can be accurately located through threshold segmentation. The experimental results demonstrate that the proposed method can effectively recognize the stem/calyx area of apple at different positions, and has good robustness and applicability with the recognition accuracy of 97.33%.
- Publication
Acta Agriculturae Zhejiangensis, 2021, Vol 33, Issue 8, p1497
- ISSN
1004-1524
- Publication type
Academic Journal
- DOI
10.3969/j.issn.1004-1524.2021.08.17