熊俊涛,林睿,刘振,何志良,杨振刚,卜榕彬.夜间自然环境下荔枝采摘机器人识别技术[J].农业机械学报,2017,48(11):28-34.
XIONG Juntao,LIN Rui,LIU Zhen,HE Zhiliang,YANG Zhen’gang,BU Rongbin.Visual Technology of Picking Robot to Detect Litchi at Nighttime under Natural Environment[J].Transactions of the Chinese Society for Agricultural Machinery,2017,48(11):28-34.
摘要点击次数: 2926
全文下载次数: 1223
夜间自然环境下荔枝采摘机器人识别技术   [下载全文]
Visual Technology of Picking Robot to Detect Litchi at Nighttime under Natural Environment   [Download Pdf][in English]
投稿时间:2017-04-07  
DOI:10.6041/j.issn.1000-1298.2017.11.004
中文关键词:  荔枝  采摘机器人  识别  夜间图像  Hough圆检测
基金项目:国家自然科学基金项目(31201135、31571568)、广东省科技计划项目(2015A020209123)和广州市科技计划项目(201506010081)
作者单位
熊俊涛 华南农业大学 
林睿 华南农业大学 
刘振 华南农业大学 
何志良 华南农业大学 
杨振刚 华南农业大学 
卜榕彬 华南农业大学 
中文摘要:利用机器视觉实现自然环境下成熟荔枝的识别,对农业采摘机器人的研究与发展具有重要意义。本文首先设计了夜间图像采集的视觉系统,然后选取了白天和夜间两种自然环境下采集荔枝图像,分析了同一串荔枝在白天自然光照与夜间LED光照下的颜色数据,确定了YIQ颜色模型进行夜间荔枝果实识别的可行性。首先选择夜间荔枝图像的I分量图,利用Otsu算法分割图像去除背景,然后使用模糊C均值聚类算法分割果实和果梗图像,得到荔枝果实图像;再利用Hough圆拟合方法检测出图像中的各个荔枝果实。荔枝识别试验结果表明:夜间荔枝图像识别的正确率为95.3%,识别算法运行的平均时间为0.46s。研究表明,该算法对夜间荔枝的识别有较好的准确性和实时性,为荔枝采摘机器人的视觉定位方法提供了技术支持。
XIONG Juntao  LIN Rui  LIU Zhen  HE Zhiliang  YANG Zhen’gang  BU Rongbin
South China Agricultural University,South China Agricultural University,South China Agricultural University,South China Agricultural University,South China Agricultural University and South China Agricultural University
Key Words:litchi  picking robot  recognition  nighttime image  Hough circle detection
Abstract:Fruit and vegetable production occupy an important position in agriculture with wide market and huge economic benefit. Currently, due to the diversity of picking object, most of fruit harvesting in our country depends on manual work. It’s not only time-consuming, but also technic-demanding. The labor cost of harvesting tends to occupy one-third to one-half of the whole labor cost in fruit production process. Thus, fruit harvesting robot needs to be developed to increase the efficiency and lower the costs. Since the working task of harvesting robot grows in natural environment with various shapes and complex structure, visual system needs to be built to recognize the target. This article focusing on litchi picking process, a visual system for litchi images was built and used to recognize litchi. Firstly, a visual system for litchi picture acquisition was built and a method of nighttime litchi recognition and picking point calculation was proposed. For comparison, pictures of same cluster of litchis were captured at daytime with different natural illumination and nighttime with artificial illumination. By analyzing color features of same litchi picture in different color models, the YIQ color model was proved to be the model with best practicability for nighttime litchi recognition and picking point calculation. The background of nighttime picture was firstly removed using Otsu algorithm, then fruit was segmented from stem using Fuzzy C-means clustering algorithm. Circle detection was performed to recognize fruits respectively using Hough circle fitting method. The experiments showed that nighttime litchi recognition accuracy was 95.3% with the average recognition time of 0.46s, and the method for litchi recognition at night time had better accuracy and higher real-time. This research provided technical support for visual localization technology of litchi picking robots. Based on machine vision, the recognition of litchi fruit was realized. It could provide technical support for litchi picking robot, bring practical significance with high harvest efficiency and low labor cost.

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