中国农业科技导报 ›› 2020, Vol. 22 ›› Issue (5): 97-105.DOI: 10.13304/j.nykjdb.2018.0501

• 智慧农业 农机装备 • 上一篇    下一篇

电动生长锥结合Python树木年轮信息提取研究

常晨1,王秀兰1*,冯仲科1,孙素芬2,张浪3,4,马文苑1   

  1. 1.北京林业大学, 精准林业北京市重点实验室, 北京 100083; 2.北京市农林科学院农业信息与经济研究所, 北京 100097; 3.上海市园林科学规划研究院, 上海 200232; 4.上海城市困难立地绿化工程技术研究中心, 上海 200232
  • 收稿日期:2018-08-22 出版日期:2020-05-15 发布日期:2018-11-26
  • 通讯作者: *通信作者 王秀兰 E-mail:wangxiulan2000@sina.com
  • 作者简介:常晨 E-ail:564703750@qq.com;
  • 基金资助:
    青年教师科学研究中长期项目(2015ZCQ-LX-01);国家自然科学基金项目(U1710123);北京市自然科学基金项目(6161001)。

Research on Extracting Tree Ring Information From Python by Electric Growth Cone

CHANG Chen1, WANG Xiulan1*, FENG Zhongke1, SUN Sufen2, ZHANG Lang3,4, Ma Wenyuan1   

  1. 1.Beijing Key Laboratory of Precision Forestry, Beijing Forestry University, Beijing 100083, China; 2.Institute of agricultural information and economy, Beijing Academy of agricultural and Forestry Sciences, Beijing 100097, China; 3.Shanghai Academy of Landscape Architecture Science and Planing, Shanghai 200232, China; 4.Shanghai Engineering Research Center of Landscaping on Challenging Urban Sites, Shanghai 200232, China
  • Received:2018-08-22 Online:2020-05-15 Published:2018-11-26

摘要: 为快速精准获得树木年龄及年轮信息,研发了一种电动生长锥结合Python视觉库(PIL)进行年轮信息自动提取的方法。电动生长锥是在传统生长锥的基础上增加动力组,以电动力代替人力进行树木树芯获取,树芯经处理后使用扫描仪扫描得出的树芯图像,利用Python视觉库中的二值化、坐标信息提取方法对树芯年轮图像信息进行提取分析,根据每点之间的像素数,运用距离公式计算得出各年轮的宽度。该方法操作简单,成本低廉且精度较高,仅需使用电动生长锥配合扫描仪及电脑即可得出树木年轮信息,适用于不具备专业仪器环境的外业条件时树木树芯年轮进行快速测定。通过实验对电动生长锥适用性范围进行了分析,其电动生长锥取样与传统手动钻取相比,树芯状态无明显差异但时间节省一倍以上,利用Python脚本进行年轮信息提取与LinTab仪器相比,测量年轮宽度值未见明显差异,其Bisa%为023%,RMSE%为187%,满足年轮信息提取基本需求,且集成程度高,只需输入图片即可得到年轮信息。

关键词: 电动生长锥, 自动识别, 便携式, 设备研制

Abstract: To obtain information on tree age and tree rings quickly and accurately, a method for the rapid determination of tree ring width and tree age was developed using the Python image library (PIL). In this method, the electric growth cone was added to the power group on the basis of the traditional growth cone. The tree core was obtained by replacing human force with electric power, and the core was scanned by scanner with the binary and coordinate information extraction method in PIL to extract and analyze the image information of the tree ring. According to the number of pixels between each point, the width of each ring was calculated using a distance formula. The reported method was simple, inexpensive and highly precise. Tree ring information was obtained by using only an electric growth cone scanner and a computer, which was suitable for the rapid determination of tree core annual rings in the absence of external conditions in a professional instrument environment. The applicability range of the electric growth cone was analyzed experimentally. The sampling results of the electric growth cone were not different from those of traditional manual drilling (Lintab), but the former method was twice as fast as the latter method. The BISA% was 023%, the RMSE% was 187%, and the basic ring information demands were satisfied. The degree of integration was high enough to obtain information about the rings.

Key words: electric growth cone, automatic identification, portable, equipment development