基于显著性检测和Grabcut算法的自然背景茶叶嫩芽图像分割
Segmentation of natural background tea bud image based on salient object detection and Grabcut
投稿时间:2020-07-27  修订日期:2020-07-27
DOI:
中文关键词: 茶叶图像分割  显著性检测  Grabcut
英文关键词: tea image segmentation  salient object detection  Grabcut
基金项目:湖北省技术创新专项重大项目(2019ABA101);湖北省科技计划项目(2019CFC890)
作者单位E-mail
毛腾跃 中南民族大学 734813320@qq.com 
张雯娟 中南民族大学 605317657@qq.com 
帖军 中南民族大学  
黄印 中南民族大学  
摘要点击次数: 144
全文下载次数: 0
中文摘要:
      人工采茶存在效率低下、人工成本高等问题,机械采茶取代人工采茶成为一种重要的采茶方式,针对机械采茶中自然背景茶叶嫩芽分割效率问题,提出了一种自然背景下的茶叶嫩芽图像分割算法。算法首先利用显著性检测算法提取出图像中的突出目标作为显著性图,再结合GrabCut算法精确的分割出突出的对象。实验结果表明,该图像分割算法对于背景复杂且目标与背景对比不明显的茶叶图像具有良好的分割效果,这为机械手采茶提供了一种新的方法。
英文摘要:
      Artificial tea picking has problems such as low efficiency and high labor cost. Mechanical tea picking has replaced artificial tea picking as an important way of tea picking. Aiming at the segmentation efficiency of natural background tea buds in mechanical tea picking, this paper proposes a natural background tea bud image segmentation algorithm. Firstly, the salient objects in the image are extracted as the saliency map by the saliency detection algorithm, and then the salient objects are accurately segmented by combining with the grabcut algorithm. The experimental results show that the image segmentation algorithm has good segmentation effect for tea images with complex background and no obvious contrast between target and background, which provides a new method for tea picking by manipulator.
View Fulltext   查看/发表评论  下载PDF阅读器
关闭