Neuro-Fuzzy-Based Smart Irrigation System and Multimodal Image Analysis in Static-Clustered Wireless Sensor Network for Marigold Crops

Neuro-Fuzzy-Based Smart Irrigation System and Multimodal Image Analysis in Static-Clustered Wireless Sensor Network for Marigold Crops

Karthick Raghunath K. M. (Malla Reddy Institute of Engineering and Technology, India) and Anantha Raman G. R. (Malla Reddy Institute of Engineering and Technology, India)
DOI: 10.4018/978-1-7998-3591-2.ch015
OnDemand PDF Download:
Available
$37.50
No Current Special Offers
TOTAL SAVINGS: $37.50

Abstract

As a decorative flower, marigolds have become one of the most attractive flowers, especially on the social and religious arena. Thus, this chapter reveals the potential positive resultants in the production of marigold through neuro-fuzzy-based smart irrigation technique in the static-clustered wireless sensor network. The entire system is sectionalized into clustering phase and operational phase. The clustering phase comprises three modules whereas the operational phase also includes three primary modules. The neuro-fuzzy term refers to a system that characterizes the structure of a fuzzy controller where the fuzzy sets and rules are adjusted using neural networks iteratively tuning techniques with input and output system data. The neuro-fuzzy system includes two distinct way of behavior. The vital concern of the system is to prevent unnecessary or unwarranted irrigation. Finally, on the utilization of multimodal image analysis and neuro-fuzzy methods, it is observed that the system reduces the overall utilization of water (~32-34%).
Chapter Preview
Top

Introduction

In India, marigold is one of the most commonly grown flowers and used extensively during religious and social functions in various forms. Based on the cultivation report (Sujitha & Shanmugasundaram, 2017), the southern state namely Tamil Nadu which ranks number one among the marigold production, that covers 6733 hectare with an estimated production range of ~100995 tonnes annually, and the productivity ranges around 15 tonnes/hectare. Most of the time, the marigold flowers are intentionally grown as decorative crops that could be marketed as loose flowers or ornamental garlands for specialty usage. It is also one of the prominent instinctive origins of xanthophylls as an artificial additive to brighten poultry skin and egg yolks (Sujitha & Shanmugasundaram, 2017). Apart from these usages, the most part of the flowers also being used efficaciously for commercial dyestuff fabrics (Sujitha & Shanmugasundaram, 2017). High necessities of marigold flower either as a extracted products or as cut flower is always peak in most of the developed and developing countries (Italy, Taiwan, South Korea, Japan, UK, Mexico, Spain, United States). Hence, exportation of marigold flower plays a crucial role in the uplift of farmer’s economy. Although the entire cultivation of flower is systemically exercised in an open field environment, still the irrigation process lacks the regulated and efficient utilization of water which is the firm demand for the raise in yield.

Smart irrigation is rightfully became a crucial practice of modern days agriculture. For majority of the cropping strategies especially in arid and semiarid regions require an efficient and systematical irrigation system to yield the maximum production, and to elevate the socio-economic condition. Beside linear-movable irrigation systems, Self-propelled centre pivot systems are commonly applied to irrigate the agri-field quite systematically; however, crucial variances in soil properties and water availableness exist all over the fields (Crow & Wood 2003). Automated irrigation systems are efficient to irrigate plants to an appropriate level for a normal growth of plants. Conventional method results in the reduction of productivity and energy.

The modern techniques of Neuro-fuzzy have found application in almost all the fields. The “Neuro-Fuzzy” term entails a kind of system characterized for a similar structure of a fuzzy controller where the fuzzy sets and rules are adjusted using neural networks tuning techniques in an iterative way with input and output system data. The system includes two distinct way of behavior. In the first phase called learning phase, it behaves like neural network that learns its internal parameter. In the second phase called the execution phase, it behaves like a fuzzy logic system.

Neural-fuzzy networks are often referred as connectionist models that are trained as neural networks, but their system is structurized as a fuzzy rule interpreter to obtain an unambiguous, crisp resultant. A Neuro-fuzzy inference system comprises a set of linguistic rules and an inference operation that are substantiated or aggregated with a connectionist framework for better adaption.

The neuro-fuzzy controller which is developed can determine the quantity of water required by plants in well defined depth using sensor nodes such as soil moisture sensor, humidity sensor and temperature sensor. Automatic irrigation scheduling system can be used to supply the water automatically based on the measurements specified through sensors. Sensors fixed in the field are used to ensure the water level in the soil. A solution to the problem of excessive supply of water can prevent damage of crops.

The combined analysis of multi-images of the same specimen adopted in various imaging modalities is typically used to acquire more selective information about the flower growth progress than is potentially analyzed from just a single modality. Thus, after the successive execution of operational phase, to analyze the growth progress of marigold flower; Non- Subsampled Contourlet Transform (NSCT) is employed for fusing multi-modal images (L. da Cunha et al., 2006).

The proposed neuro-fuzzy based smart irrigation technique in precision agriculture of marigold cultivation is addressed in Section 2 of this chapter. The complete evaluation of smart irrigation system based upon the multimodal image analyzing is also discussed in Section 3. Justifiable and reasonable conclusion and the future research work of this proposed technique is discoursed in the Section 4 of this chapter.

Complete Chapter List

Search this Book:
Reset