Estimating Morphological Features of Plant Growth Using Machine Vision

Estimating Morphological Features of Plant Growth Using Machine Vision

Himanshu Gupta (National institute of Technology Jalandhar, Jalandhar, India) and Roop Pahuja (National Institute of Technology Jalandhar, Jalandhar, India)
DOI: 10.4018/IJAEIS.2019070103
OnDemand PDF Download:
No Current Special Offers


Motivated by the fact that human visionary intelligence plays a vital role in guiding many of the agriculture practices, this article represents an effective use of machine vision technology for estimating plant morphological features to ascertain its growth and health conditions. An alternative to traditional, manual and time-consuming testing methods of plant growth parameters, a novel online plant vision system is proposed and developed on the platform of virtual instrumentation. Deployed in real time, the system acquires plant images using digital camera and communicates the raw image to host PC on Wi-Fi network. The dedicated application software with plant user interface, effective image processing and analysis algorithms, loads the plant images, extracts and estimates certain morphological features of the plant such as plant height, leaf area, detection of flower onset and fall foliage. The system was tested and validated under real-time conditions using different plants and leaves. Further, the performance of the system was statistically analysed to show promising results.
Article Preview

1. Introduction

The human visionary and cognitive power to see, think, analyse and promptly take actions have motivated and accelerated research in the area of machine/computer vision systems over the last 50 years (Davies, 2012). Machine vision technology uses specialized devices such as image sensors/cameras, image processing software tools and/or actuators to automatically capture, process and interpret a digital image of a real scene in order to extract useful information for monitoring, decision making, classifying and/or control of machines such as manufacturing robots or processes (Rosenfeld, 1985, Lewis, 2014). Machine vision is slightly different from computer-vision that concentrates on processing and analysis of digital image. Machine vision is a combination of image acquisition hardware, processing software and optionally the use of actuators for control of process/devices to perform image-based application specific tasks accurately, repeatedly and timely with visualization of information (Szelisk, 2010). With the advancements in affordable, sophisticated image sensors, smart camera and optics with digital interfacing, image processing methods and tools, the machine vision technology is fast evolving to solve complex application tasks (Teledyne, 2014). The high-end aim of machine vision technology is to design systems and applications that can compete with the human visionary capabilities and perform tasks equivalent or better than humans in many cases (Teledyne, 2014). For e.g. computer vision system is effectively used in high speed production lines at a factory floor for detection of faults in manufactured components and sorting of materials. The work with vision system is done at much faster pace and accurately than the labour involved in the process (Lewis, 2014). In the case of medical imaging, machine-vision systems are a tool in the hand of doctors to obtain images of inner organs of the body, extract features and analyses images to provide better diagnosis, prediction and treatment of diseases (Chen, 2013).

Tools of machine vision have been extensively researched and used in broad range of applications for different tasks such as identification or recognition of objects, sorting and inspection of materials, health diagnosis, geographical and environment assessment etc. (Solari, Chessa & Sabatini, 2012). Primary focus has been in the area of industrial automation and manufacturing for material inspection, electronic components inspection and guiding of robots for control operations using 2D or more recent 3D system (Tsarouchi et al., 2016), in medical image analysis and diagnosis (Li, Ma, Wang & Zhang, 2014), (Chen, 2013). Image processing with pattern matching techniques have been used in various applications such as character/handwriting recognition (Katungunya, Ding & Mashenene, 2016), signature verification, sorting of materials based on different image attributes in food or production industry for quality assurance and analysis of satellite-based sensor images (Picon et al., 2012), (Key Technologies, 2012). Slowly and gradually, machine vision technology is spreading into other areas such as vehicle guidance or improve visibility during night and foggy weather, traffic monitoring and control (Li et al., 2018), (Singh, Vishnu & Mohan, 2016). Machine vision technology is used for inventory control and management such as barcode reading, counting of articles etc., currency identification, defence, security and surveillance (Kar, Shrikhande & Babu, 2016), crime investigation (Inspect, 2010), biometric measurement for personal identification, detection of diseases (Kallen, 2016), habitat monitoring of animals and birds, monitoring of agriculture fields and pest identification (Vázquez-Arellano, Griepentrog, Reiser & Paraforos, 2016).

Complete Article List

Search this Journal:
Open Access Articles
Volume 12: 4 Issues (2021): Forthcoming, Available for Pre-Order
Volume 11: 4 Issues (2020)
Volume 10: 4 Issues (2019)
Volume 9: 4 Issues (2018)
Volume 8: 4 Issues (2017)
Volume 7: 4 Issues (2016)
Volume 6: 4 Issues (2015)
Volume 5: 4 Issues (2014)
Volume 4: 4 Issues (2013)
Volume 3: 2 Issues (2012)
Volume 2: 2 Issues (2011)
Volume 1: 2 Issues (2010)
View Complete Journal Contents Listing