基于人工智能的苗木质量无损检测研究进展  

Research Progress in Non-destructive Testing of Seedling Quality Based on Artificial Intelligence

作　　者：李玉荣[1] 刘英[1] 汪力 费叶琦[1] Li Yurong;Liu Ying;Wang Li;Fei Yeqi(College of Mechanical and Electronic Engineering,Nanjing Forestry University,Nanjing 210037,China)

机构地区：[1]南京林业大学机械电子工程学院,南京210037

出　　处：《世界林业研究》2020年第6期27-32,共6页World Forestry Research

基　　金：国家自然科学基金面上项目“基于多信息融合的马尾松苗木活力快速精准预测模型研究”(31570714);江苏省重点研发计划(产业前瞻与关键核心技术)项目(BE2019112)。

摘　　要：我国森林面积广域,人工林面积居世界首位。为了确保质优量足的苗木进一步提升苗木造林效果,苗木质量的无损检测成为苗木质量精准快速评价的关键。文中概述人工智能理论与算法在苗木质量的形态、生理和活力指标3个方面的无损检测应用现状;针对传统检测指标单一、效率低和主观误差大的问题,指出综合应用图像采集、数字图像处理和机器学习技术的人工智能算法与理论在苗木质量评价指标检测领域具有明显优势,并从检测技术融合、提升检测算法和多源数据融合等方面进行展望,旨在为苗木质量快速精准评价提供参考。China boasts large forest area, and her artificial forest tops the world in area. In order to ensure the quality and quantity of seedling production and further improve the effect of afforestation with seedlings, the non-destructive testing of seedling is the key to the accurate and rapid evaluation of seedling quality. This paper reviews the application of artificial intelligent theory and algorithm to the non-destructive testing of morphological index, physiological index and vitality index of seedlings. In order to solve the problems of the traditional testing including unitary index, low efficiency and more subjective errors, the artificial intelligent theory and algorithm that integrates image acquisition, digital image processing and machine learning technology has obvious advantages in the field of seedling quality evaluation index testing. The paper also prospects the testing technology fusion, improvement of the testing algorithm and multi-source data fusion, with the expectation to provide references for rapid and accurate evaluation of seedling quality.

关 键 词：苗木质量 数字图像处理 无损检测 机器学习 

分 类 号：S723[农业科学—林木遗传育种]