الفهرس 
IntroductionOnly 14 pages are availabe for public view
 |  | from | 105 |  |  |
| | | from | 105 | | |
Abstract
Nowadays, the need of smart cities is growing and one of its most important industries is precise farming in which sensors need to be used in increasing the yield in both quality and quantity and preventing pests and threatening diseases. This research is proposing a design of a system to early detect infected palm trees with Red Palm Weevil (RPW). This can be done using artificial intelligence inferred on FPGA. FPGA provides the ability of data sampling and processing due to its flexibility and low power consumption in AI field. This proposal includes the usage of DenseNet which is an artificial neural network on an FPGA platform to classify the audio signal collected from a selected group of palm trees in a given field after converting it into spectrogram using MFCC techniques. DenseNet CNN has been used, trained on PC then inferred and tested on FPGA.
The presented thesis consists of four main chapters which are summarized as follows:
Chapter 1 This is the Introduction chapter in which the importance of the dates cultivation is introduced and the problem encountering this field is stated.
Chapter 2 This is the Literature Survey chapter in which a survey is presented of different early detection techniques of the RPW infestation and different machine learning techniques along with the FPGA usage in AI applications.
Chapter 3 This is the Methodology chapter in which the proposed methodology is explained in details and the results are presented.
Chapter 4 This is the Conclusion and Future Work chapter in which the final conclusion of this thesis is presented and the possibility of using this work as a start for a real time detection.