# Scene and Environment Monitoring Using Aerial Imagery and Deep Learning

**Authors:** Mahdi Maktabdar Oghaz, Manzoor Razaak, Hamideh Kerdegari, Vasileios Argyriou, Paolo Remagnino  
**arXiv:** [1906.02809](https://arxiv.org/abs/1906.02809)  
**Date:** 2019-06-06  
**Summary page:** https://www.summarizepaper.com/en/paper/1906.02809v1/

## Key points

- Unmanned Aerial Vehicles (UAVs) are promising for smart farming applications, especially crop monitoring
- Aerial imagery captured by UAVs facilitates decision-making processes in agriculture
- Deep learning techniques improve precision and reliability of analyzing aerial imagery
- Various sensors can be mounted on UAVs for remote crop analysis tasks
- Reviewed studies apply deep learning algorithms to UAV imagery for smart farming purposes
- Studies categorized into vegetation identification, classification and segmentation; crop counting and yield predictions; crop mapping; weed detection; and detection of crop diseases and nutrient deficiencies
- Paper explores potential of using UAVs equipped with deep learning algorithms for monitoring scenes and environments in agriculture farms
- Versatility of UAVs in terms of sensor mounting options discussed
- Numerous studies have been conducted on the use of UAVs in smart farming applications

## Summary

Unmanned Aerial Vehicles (UAVs) have emerged as a promising technology for smart farming applications, particularly in the field of crop monitoring. By utilizing aerial imagery captured by UAVs, key decision-making processes related to agriculture can be facilitated. Recent advancements in deep learning techniques have further improved the precision and reliability of analyzing aerial imagery. The ability to mount various sensors, such as RGB and spectral cameras, on UAVs allows for remote crop analysis tasks including vegetation classification and segmentation, crop counting, yield monitoring and prediction, crop mapping, weed detection, and identification of crop diseases and nutrient deficiencies. This paper presents a comprehensive review of studies that apply deep learning algorithms to UAV imagery for smart farming purposes. The reviewed studies are categorized into five major groups based on their specific applications: vegetation identification, classification and segmentation; crop counting and yield predictions; crop mapping; weed detection; and detection of crop diseases and nutrient deficiencies. Each study is critically analyzed in depth. In addition to the existing summary, this paper titled "Scene and Environment Monitoring Using Aerial Imagery and Deep Learning" by Mahdi Maktabdar Oghaz et al., explores the potential of using UAVs equipped with deep learning algorithms for monitoring scenes and environments. The authors highlight how UAVs can be employed in agriculture farms to make informed decisions regarding crops through aerial monitoring. They emphasize that advancements in deep learning techniques have significantly enhanced the accuracy and dependability of analyzing aerial images for agricultural purposes. The authors also discuss the versatility of UAVs in terms of sensor mounting options. By attaching different types of sensors like RGB cameras or spectral cameras onto UAVs, farmers can remotely analyze crops for various purposes such as identifying vegetation types, classifying crops accurately, estimating yields precisely, creating detailed maps of cultivated areas, detecting weeds effectively as well as identifying diseases or nutrient deficiencies affecting crops. Furthermore, the paper acknowledges that numerous studies have been conducted exploring the use of UAVs in smart farming applications. To provide a comprehensive overview, the authors classify these studies into five main categories based on their specific focus areas: vegetation identification; classification and segmentation; crop counting and yield predictions; crop mapping; weed detection; and detection of crop diseases and nutrient deficiencies. The paper then proceeds to critically analyze each study in detail. Overall, this paper contributes to the existing body of knowledge by presenting a thorough review of studies that utilize deep learning algorithms on UAV imagery for smart farming purposes. It highlights the potential benefits of using UAVs equipped with advanced image analysis techniques in agriculture providing valuable insights for researchers and practitioners in the field.