|
" Crop Yield Prediction Using Satellite Remote Sensing and Artificial Neural Network "
Akhand, Kawsar
Roytman, Leonid
| Document Type
|
:
|
Latin Dissertation
|
| Language of Document
|
:
|
English
|
| Record Number
|
:
|
1051022
|
| Doc. No
|
:
|
TL50139
|
| Main Entry
|
:
|
Akhand, Kawsar
|
| Title & Author
|
:
|
Crop Yield Prediction Using Satellite Remote Sensing and Artificial Neural Network\ Akhand, KawsarRoytman, Leonid
|
| College
|
:
|
The City College of New York
|
| Date
|
:
|
2019
|
| Degree
|
:
|
Ph.D.
|
| student score
|
:
|
2019
|
| Note
|
:
|
217 p.
|
| Abstract
|
:
|
Agriculture is a strategic sector that plays a crucial role in the growth of the economy, employment generation and poverty reduction all over the world. The main goal of the world’s agriculture and grain sector is to provide food for 7.6 billion people. Bangladesh is essentially an agro-based developing country where people’s livelihood and economic progress depend on agriculture. Due to the increase in population and the decrease of agricultural land, this sector is under pressure to ensure food security for its vast population. The continuous evolution of technological advancement helps to maximize the crop production, improvise the quality of crop and to ensure food safety. While the modernization in the agricultural sector is still in progress, there is another area of interest which is the quantitative, effective and timely estimation of crop yield before harvest in the different geographic locations. Early prediction of crop yield can provide valuable information for the government and its stakeholders to maintain food security, minimize risk associated with production, prices, reservation and trade. The weather conditions of an area such as temperature, humidity, precipitation, sunshine and atmospheric carbon dioxide play an important role in determining its agricultural production. Satellite remote sensing is a potential technology to provide adequate spatial and temporal information of environment at a global scale in a cost-effective, regular, accurate and timely manner that can be used for crop growth, monitoring and yield estimation. Satellite-derived vegetation health indices have strong correlations with crop vegetation health, conditions and productivity. This research investigates the application of the Advanced Very High Resolution Radiometer (AVHRR) sensor-based vegetation health indices Vegetation Condition Index (VCI) characterizing moisture condition and Temperature Condition Index (TCI) characterizing thermal condition as proxies for crop yield prediction before harvest. Artificial Neural Network (ANN) is widely used as a crop yield prediction application because of its ability to learn complicated and nonlinear relationships between the different parameters of crop growth and yield. This study demonstrates the successful application of ANN and remote sensing satellite data to develop a reliable and efficient crop yield prediction model. The proposed model is applied to predict the yield of four main crops (Boro Rice, Aus Rice, Wheat and Potato) cultivated in different seasons and at different time of the year, in Bangladesh. The performance of this model shows great promise in prediction capabilities.
|
| Descriptor
|
:
|
Remote sensing
|
| Added Entry
|
:
|
Roytman, Leonid
|
| Added Entry
|
:
|
The City College of New York
|
| |