刘刚,张雪,宗泽,郭彩玲.基于深度信息的草莓三维重建技术[J].农业机械学报,2017,48(4):160-165,172.
LIU Gang,ZHANG Xue,ZONG Ze,GUO Cailing.3D Reconstruction of Strawberry Based on Depth Information[J].Transactions of the Chinese Society for Agricultural Machinery,2017,48(4):160-165,172.
摘要点击次数: 2145
全文下载次数: 1376
基于深度信息的草莓三维重建技术   [下载全文]
3D Reconstruction of Strawberry Based on Depth Information   [Download Pdf][in English]
投稿时间:2016-07-26  
DOI:10.6041/j.issn.1000-1298.2017.04.021
中文关键词:  草莓冠层  三维重建  深度信息  点云聚类
基金项目:国家高技术研究发展计划(863计划)项目(2013AA102406)
作者单位
刘刚 中国农业大学 
张雪 中国农业大学 
宗泽 中国农业大学 
郭彩玲 中国农业大学 
中文摘要:以盆栽和高架两种栽培模式生长环境下的草莓植株为研究对象,提出了一种基于深度信息分割聚类的草莓冠层结构形态三维重建算法。首先,以深度信息的不连续性特征作为草莓植株逐层分割的重要依据,以深度二维图像作为全局参考指标,提出深度信息步进方法,自动提取冠层点云;其次,改进密度聚类算法,有效滤除随机、跳边和背景噪声;最后,改进基于Harris算子的多源图像融合算法,实现彩色图像与强度图像的配准及点云颜色的映射,三维重建出具有颜色信息的草莓冠层结构形态。为验证该算法的有效性,将三维重建后冠层的平均单叶长度及A-B线距离作为评价指标,试验结果表明,模型的平均单叶长度计算正确率为93%左右,A-B线距离计算正确率为97%左右,研究结果可为草莓采摘机器人果实识别过程中枝叶空间结构关系的构建提供技术支持。
LIU Gang  ZHANG Xue  ZONG Ze  GUO Cailing
China Agricultural University,China Agricultural University,China Agricultural University and China Agricultural University
Key Words:strawberry canopy  3D reconstruction  depth information  point cloud clustering
Abstract:With the development of 3D acquisition devices, 3D modeling method emerges in endlessly. In recent years, 3D reconstruction based on real data is focused on scanning and TOF imaging method. The main drawback of the former is that needs extensive man-machine interaction. The latter can transform a wide range of scene depth information into plant 3D point cloud data in a short time. Most plant reconstruction method based on TOF technology is center on brance structure, but for herbaceous plant, leaves are not easy to operate. So a 3D model reconstruction algorithm was presented based on depth information segmentation and clustering for potted and elevated strawberry. Firstly, the algorithm took discrete depth information as a significant reference standard for object segmentation while taking the depth-dimensional image as a global reference standard to extract discrete point cloud through depth stepping. Then, the algorithm used clustering algorithm based on density to filter out random noise, jump-edge noise as well as background noise. By applying Harris algorithm between color image and the intensity image, a robust registration result was got and an index relationship between point cloud and pixel points was found. Ultimately, 3D strawberry canopy morphology with full color information was reconstructed after following all those steps. Experimental results showed that the algorithm had achieved good effects on segmentation clustering and coloring for potted and elevated strawberry leaves. The accuracy of single-leaf length was around 93% while A-B line accuracy was close to 97%, which indicated that the reconstructed model accuracy met the requirements of phenotypic parameters extraction. The 3D model can provide a new strategy for picking robot in spatial structure research domain.

Transactions of the Chinese Society for Agriculture Machinery (CSAM), in charged of China Association for Science and Technology (CAST), sponsored by CSAM and Chinese Academy of Agricultural Mechanization Science(CAAMS), started publication in 1957. It is the earliest interdisciplinary journal in Chinese which combines agricultural and engineering. It always closely grasps the development direction of agriculture engineering disciplines and the published papers represent the highest academic level of agriculture engineering in China. Currently, nearly 8,000 papers have been already published. There are around 3,000 papers contributed to the journal each year, but only around 600 of them will be accepted. Transactions of CSAM focuses on a wide range of agricultural machinery, irrigation, electronics, robotics, agro-products engineering, biological energy, agricultural structures and environment and more. Subjects in Transactions of the CSAM have been embodied by many internationally well-known index systems, such as: EI Compendex, CA, CSA, etc.

   下载PDF阅读器