引用本文: | 徐树奎,周浩.对象边框标注数据的弱监督图像语义分割[J].国防科技大学学报,2020,42(1):187-193.[点击复制] |
XU Shukui,ZHOU Hao.Image semantic segmentation of weakly supervised model with bounding box annotations data[J].Journal of National University of Defense Technology,2020,42(1):187-193[点击复制] |
|
|
|
本文已被:浏览 6075次 下载 5477次 |
对象边框标注数据的弱监督图像语义分割 |
徐树奎1,周浩2 |
(1. 中国电子科技集团第28.研究所 博士后流动工作站, 江苏 南京 210007;2. 国防科技大学 系统工程学院, 湖南 长沙 410073)
|
摘要: |
针对图像语义分割应用中像素级标注数据费时昂贵的问题,主要研究以对象边框标注数据为代表的弱监督模型下的图像语义分割方法。使用基于金字塔的密集采样全卷积网络提取图像的像素级特征,并用GrabCut算法转化对弱监督数据进行数据标记,通过将图像特征和标记数据进行联合训练,构建了基于金字塔密集采样全卷积网络的对象边框标注弱监督图像语义分割模型,并在公开数据集上进行了验证。实验结果表明,所构建的弱监督模型与DET3-Proposed模型、全矩形转化模型以及Bbox-Seg模型相比,达到了更好的分割效果。 |
关键词: 图像语义分割 全卷积网络 弱监督模型 GrabCut算法 |
DOI:10.11887/j.cn.202001026 |
投稿日期:2018-11-01 |
基金项目:国家自然科学基金资助项目(61671459) |
|
Image semantic segmentation of weakly supervised model with bounding box annotations data |
XU Shukui1, ZHOU Hao2 |
(1. Mobile Postdoctoral Center, The 28.th Research Institute of China Electronic Technology Group Corporation, Nanjing 210007, China;2. College of Systems Engineering, National University of Defense Technology, Changsha 410073, China)
|
Abstract: |
For the pixel level marked data of image segmentation is time consuming and expensive, the image segmentation under the weak supervised model with bounding box annotations was studied. The high-resolution pixel features were extracted by a pyramid densely sampled fully convolution network, and then the weakly supervised data by the GrabCut algorithm was transformed. Finally, the image features and the labeled data were jointly trained. A model of weakly supervised image semantic segmentation based on fully convolution network was constructed and verified on public data set. Experiments show that the constructed weak supervised model has a better segmentation effect compared with DET3-Proposed model, expectation-maximization model and Bbox-Seg model. |
Keywords: image semantic segmentation fully convolution network weakly supervised model GrabCut algorithm |
|
|
|
|
|