AI Communication With voice Command Execution, Assignments of Artificial Intelligence

Download AI Communication With voice Command Execution and more Assignments Artificial Intelligence in PDF only on Docsity!

Use of Remote sensing for cotton crop

identification and monitoring the condition

By

Maham Asif Bukhari

CIIT/FA16-BES-018 /VHR

BS Thesis

COMSATS University Islamabad

Vehari Campus- Pakistan

Spring, 2020

i

COMSATS University Islamabad, Vehari Campus

Thesis Title

Use of Remote sensing for cotton crop identification

and monitoring the condition

A Thesis Presented to

COMSATS University Islamabad, Vehari Campus

In partial fulfillment

of the requirement for the degree of

Bachelor of Science

in

Environmental Sciences

By

Maham Asif Bukhari

CIIT/FA16-BES-018/VHR

Spring, 2020

ii

Final Approval

This thesis titled

Use of Remote Sensing for cotton crop identification and monitoring the condition

By

Maham Asif Bukhari

CIIT/FA16-BES-018/VHR

Has been approved

For the COMSATS University Islamabad, Vehari Campus

External Examiner: _________________________________________________ Dr. Tanveer ul Haq Department of soil and Environmental Sciences MNS University of Agriculture, Multan Supervisor: _________________________________________________________ Dr. Muhammad Mubeen Assistant Professor Department of Environmental Sciences, CUI Vehari Member Supervisory Committee: _______________________________________ Dr. Wajid Nasim Associate Professor Department of Agronomy, IUB Bahawalpur Member Supervisory Committee: _______________________________________ Dr. Hafiz Mohkum Hammad Assistant Professor Department of Environmental Sciences, CUI iv

Vehari Declaration I, Maham Asif Bukhari (CIIT/FA16-BES-018/VHR) hereby declare that I have produced the work presented in this thesis, during the scheduled period of study. I also declare that I have not taken any material from any source except referred to wherever due. If a violation of HEC rules on research has occurred in this thesis, I shall be liable to punishable action under the plagiarism rules of the HEC. Date: _Thesis submission date Signature of the student: __________________ (Maham Asif Bukhari) (CIIT/FA16-BES-018/VHR) v

DEDICATION To Almighty ALLAH and the Holy Prophet Muhammad (P.B.U.H) & My Loving Family vii

ACKNOWLEDGEMENTS

In the name of ALLAH , the beneficent, the merciful. O my Lord! Relieve my mind and easy my task for me; and loose a knot from my tongue. That they may understand what I say (AL- Quran). I pay my sincere and humble thanks to Almighty ALLAH, the most kind, Who has helped me to overcome my problems and worries. I am thankful to ALLAH that He sent Prophet MOHAMMAD (S.A.W) for our guidance who is the most perfect and best of all ever born on the earth surface. I sincerely pay my humble gratitude from the bottom of my heart to my supervisor, Dr. Muhammad Mubeen for providing his assistance, guidance and kindness during the course of the research. I would like to pay my most humble regard to my co-supervisor Dr. Mohkum Hammad for his sincere guidance, patience, hard work and encouragement. I wish to regard my sincere gratitude to my Head of Department, Dr. Muhammad Shahid for their dexterous guidance and supervision during this research. At the end, I would pay my heartily gratitude to my friends specially Miss Tasmiya Jabeen and Rimsha Naeem and my research fellows Miss Maham Fatima and Mr. Muhammad Sani Ud Din for their cooperation and assistance during this research. I must acknowledge the debt of gratitude which I owe to my loving family (especially Mr. Syed Asif Bukhari and Miss Humaira Bukhari) for all that I have been able to achieve. Without their collaboration, love, encouragement and help I would not have been able to complete this thesis. Maham Asif Bukhari CIIT/FA16-BES-018/VHR viii

TABLE OF CONTENTS

4. Discussion…………………………………………………………….......

5. Conclusion……………………………………………………………….

6. References………………………………………….……………………

xi

LIST OF TABLES

Table 2.1 Landsat 8 dataset ……………………………...………….……. Table 2.2 Bands of Landsat 8 ……………………………………………… Table 2.3.1, Area of classified image…………………………………………. Table 2.4 NDVI max. and min. values of 2014, 2016 and 2019 years……. xiii

LIST OF ABBREVIATIONS

RS Remote Sensing GIS Geographic Information System GDP Gross Domestic Product NOAA National Oceanic and Atmospheric Administration AVHRR Advanced Very High-Resolution Radiometer MODIS Moderate Resolution Imaging Spectroradiometer NDVI Normalized Difference Vegetation Index TM Thematic Mapper GEOGLAM Global Agricultural Monitoring MARS Agricultural Resources Monitoring system SAR Synthetic Aperture Radar PCA Principle Component Analysis DL Deep Learning CNNs Coevolutionary NNs VCI Vegetation Condition Index VPMs Vegetation Phenology Metrics NIR Near Infrared USGS U.S. Geological Survey xiv

Chapter 1 Introduction 1

1. Introduction: Agriculture is the backbone of Pakistan's economy, providing employment and roughly contributing to 67 percent of the population and for Gross National Product, 35 per cent. The main cash crop is cotton. Cotton growing earns the largest export income of country and, besides lint, seeds of cotton are used for meal and oil that gives 80% of national oilseed production. Cotton and cotton products contribute 10 percent of the real gross domestic product ( GDP) of country and 55 percent of the country's foreign exchange earnings(Rehman et al. 2019). Up-to - date data is required for a proficient production and management of cotton crops. Early detection of crops plays a vital role and creates a difference in finding real estate, yielding crops on the off chance that help goes deep it helps in detecting stress, identifying crop diseases (Kharat et al. 2015). Remote sensing technology in recent years has proved to be increasingly useful in gathering information for viable resource management and thus can also be applied to monitoring and management of the coastal climate. In addition, the application of the Geographical Information System ( GIS) in the analysis of patterns and the evaluation of changes that have occurred in different subjects helps in the decision-making process of management (Sajjad and Kumar 2019). Remote sensing methods are more efficient to conventional methods, because they are quick and economical. Remote sensing applications are crop inventories, crop classification, crop production predictions, drought and flood damage assessment (Kharat et al. 2015).. So, it is important to keep track of agricultural data, remote sensing systems with their descriptive viewing capacity and variety of spatial and temporal resolution aid in the same. Remote sensing strategies are common to traditional strategies, since they are fast and financial. (Musande et al. 2012). We have mainly emphasized the use of multispectral temporal data on crop classification. Crop diversity, crop phenology, field size, and soil quality play an significant role in crop classification geographic area (Kharat et al. 2015). Plant researchers and influential persons are observing for new approaches to increase the use sustainable agricultural practices, given the fact that sustainable agriculture may be a complicated development and is still under way. The change to sustainable agriculture forces different risks on farmers compared to other sorts 2

Remote sensing information is widely used for monitoring the national and local crop distribution (Shi et al. 2013) and cropping intensity (Biradar and Xiao 2011; Son et al. 2014)of crop at local and provincial scales. Advanced Very High-Resolution Radiometer (AVHRR) to distinguish crop types; Moderate Resolution Imaging Spectroradiometer (MODIS) sensors images also were required for mapping crop types (Doraiswamy and Cook 1995); Moderate Resolution Imaging Spectroradiometer (MODIS) sensors images also were required for mapping crop types. MODIS and AVHRR sensors have daily revisit frequency, may provide continuous, global images for crop monitoring, but their spatial resolutions are generally coarser than most of crop field sizes. Mixed pixel presence has an effect on the the accuracy of mapping of ground Cover. Landsat TM / ETM+/OLI images which has fine spatial resolution, mostly used for mapping crop fields (Yang 2007). used optimized combination of three-band and NDVI from images of Landsat, and applied them to neural network models, SVM, MLC, k-nearest neighbor method, multiple classifier combinations and minimum distance method for mapping paddy rice. extracted arable land area using the information acquires from Landsat TM in the dry season (Wang et al.). Cotton is one of the most vital global economic crops, and the timely and efficient monitoring and mapping of cotton production is therefore very critical for agricultural sustainability, successful policy management and decision-making. The objective of this study is to create cotton’s spatial distribution map at spatial resolution of 30m, using Landsat8 OLI multi-temporal images in 2013. In this study, multi-temporal Landsat8 OLI images were generated by NDVI time series, then we used an effective hierarchical classification algorithm which was based on NDVI time series and reflectance images to map cotton planting area (Wang et al.). Traditional monitoring of crop growth is carried out through laboratory tests and field investigations analyzing the characteristics of growth in relation to population density, phenological development, leaf area index and aboveground biomass. At a regional level, these strategies require time, are laborious and difficult to enforce. The advantage of Remote sensing is that it is fast compared to conventional methods, and this is a world- wide technology which can be used to monitor growth of crop (Courault et al. 2016; Khanal et al. 2017; Mulla 2013; Wu et al. 2014). Several remote sensing based 4

agricultural monitoring systems have previously been established at state, national, and global scales. Global Agricultural Monitoring (GEOGLAM) has been established by the Earth Observations Group to coordinate monitoring systems of satellite in different regions of the world in order to offer coarse (national) crop production forecasts. The European Union's Agricultural Resources Monitoring (MARS) system incorporates remote sensing facts, meteorological information and predictions, and existing maps and statistics to provide information that defines, recognizes as well as analyses the key production areas of crop in Europe with a view to providing projections initially in the year and testing the farmers' subsidy application’s viability. Normalized Vegetation Difference Index (NDVI) has been used as crop monitoring system which is the major vegetative indicator of condition of crop. NDVI values are correlated and evaluated with other factors, like the nitrogen content and biomass of crops above ground, area of the root, light detection and usage. (Esquerdo et al. 2011; Meng et al. 2013; Zhou et al. 2018). The NDVI is therefore used for processes monitoring and also in real-time monitoring. Process monitoring contrasts NDVI improvements over time with those relating to the last year, (Wu et al. 2014), , while real-time monitoring assesses the variance between the multi-year average and the present year (Huang et al. 2010). Both of the approaches give the NDVI di fference (dNDVI) understood in the framework oference (dNDVI) understood in the framework of growth state of crop. (Li et al. 2019) Objectives of this study is classifying the cotton crop based on multispectral temporal data. This can be done by using Landsat data. The second objective is monitoring the crop condition using the vegetative indices i.e. NDVI. 1.1 Role of Remote sensing in agriculture: In the initial stages of remote satellite sensing, the focus of most researchers was on using information for groundcover types classification, with types of crop being a point of attention for those who had interest in agricultural applications. In recent years the research remote sensing’s role in the agriculture has centered mainly on characterizing biophysical properties of plants. Remote sensing is mostly used for analyzing and monitoring of agricultural activities over a long time (Shanmugapriya et al. 2019). 5