Crop Nutrition and Computer Vision Technology
- PDF / 755,270 Bytes
- 13 Pages / 439.37 x 666.142 pts Page_size
- 4 Downloads / 261 Views
Crop Nutrition and Computer Vision Technology Qiu Peng1 · Weihong Xu1
© Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract Crop nutrition status can be reflected via leaf state and surface color. Traditional crop nutrition testing is mainly dominated by expert experience and chemical measurement, and expert experience is greatly affected by subjective factors. Although chemical measurement has high detection accuracy, its poor timeliness makes it difficult to achieve dynamic feedback control of nutrient solution, and the sampling process will bring certain damage to the crop. Computer vision technology such as hyperspectral remote sensing, visual image and 3D scanning detection technology has become a hotspot of crop nutrition detection because of its non-destructive, rapid and real-time characteristics. It is expected to develop into the main technology for real-time diagnosis of crop nutrition, provide basis for online information monitoring of crop nutrition and timely fertilization, thereby achieving automated and intelligent management of agricultural production. Keywords Nitrogen, phosphorus and potassium · Hyperspectral remote sensing · Visual image · 3D scanning detection · Diagnostic model
1 Introduction Common methods for distinguishing nutrient elements in crops include morphological diagnosis, chemical diagnosis, fertilization diagnosis and enzymology diagnosis. These traditional diagnostic methods are time consuming and strenuous, subject to experimental environments or climatic conditions, and also not suitable for agricultural production practice. With the widespread use of computers and continuous decline in price of image processing equipment, computer vision technology has wider application in agricultural engineering. One hotspot of agricultural technology is to monitor crop growth in the field as a support to agricultural production using non-invasive monitoring technology and measurement methods. Computer vision technology is a technology that adopts image sensors for monitoring and judgment in replace of human eyes. It acquires image of the target * Weihong Xu [email protected] Qiu Peng [email protected] 1
College of Resources and Environmental Sciences, Southwest University, Chongqing 400715, People’s Republic of China
13
Vol.:(0123456789)
Q. Peng, W. Xu
object through optical devices and non-contact sensors, and then processes the image of the obtained target object by image processing software. To extract information such as color and texture and make various calculations and judgment represents an interdisciplinary subject involving multiple fields like computer science, image processing, artificial intelligence and pattern recognition. Compared with human visual system, machine vision technology boasts many advantages, making computer vision technology enjoy broad prospects in agricultural field. Thomas et al. [1] studied leaf spectral characteristics of 7 plants at different nitrogen nutrient levels, finding that reflectance of visible light band
Data Loading...
