Jadwal Sholat

Memuat jadwal sholatโ€ฆ

Pertanian & Lingkungan editorial

Open AccessOA2024

Computer Vision in Smart Agriculture and Precision Farming: Techniques and Applications

This survey explores how computer vision and AI technologies are transforming precision agriculture by enabling tasks such as crop monitoring, yield prediction, and automated treatment. It highlights key imaging techniques, the digital life cycle of crops, and the challenges of deploying generalized vision models for fully autonomous farms.
Sumaira Ghazal; Arslan Munir; W. S. Qureshiยท Artificial Intelligence in Agricultureยท 2024ยท DOI 10.1016/j.aiia.2024.06.004

The core problem

The integration of digitization and automation into agriculture has sparked a revolution driven by computer vision and artificial intelligence (AI). This transformation promises increased productivity and economic growth while addressing global issues such as food security and sustainability. Precision agriculture relies on vision-based intelligent systems to monitor crops, optimize inputs, and automate operations. This survey provides a holistic understanding of the integration of such systems across various aspects of precision farming. It focuses on widely used imaging and image analysis techniques for tasks in the digital life cycle of crops, including image acquisition, stitching and photogrammetry, image analysis, decision making, treatment, and planning. The paper also discusses salient crop metrics used in digital agriculture and concludes by outlining challenges for real-time deployment of fully autonomous farms.

Innovation

As a survey paper, the methodology involves a systematic review of existing literature on computer vision techniques applied to precision agriculture. The authors organize the review around the digital life cycle of crops, covering key phases: image acquisition, image stitching and photogrammetry, image analysis, decision making, treatment, and planning. They examine widely used imaging modalities (e.g., RGB, multispectral, hyperspectral, thermal) and image analysis methods (e.g., deep learning for classification, detection, segmentation). The survey also identifies crop metrics relevant to digital agriculture, such as plant height, leaf area index, and yield components. By structuring the discussion along the crop life cycle, the paper provides a comprehensive overview of how vision-guided systems are implemented in challenging agricultural environments.
Introduction
The integration of digitization and automation into agriculture has sparked a revolution driven by computer vision and artificial intelligence (AI). This transformation promises increased productivity and economic growth while addressing global issues such as food security and sustainability. Precision agriculture relies on vision-based intelligent systems to monitor crops, optimize inputs, and automate operations. This survey provides a holistic understanding of the integration of such systems across various aspects of precision farming. It focuses on widely used imaging and image analysis techniques for tasks in the digital life cycle of crops, including image acquisition, stitching and photogrammetry, image analysis, decision making, treatment, and planning. The paper also discusses salient crop metrics used in digital agriculture and concludes by outlining challenges for real-time deployment of fully autonomous farms.

Why it matters

The authors discuss several challenges associated with implementing generalized computer vision models for real-time deployment in fully autonomous farms. These include:

- **Environmental variability**: Changes in lighting, weather, and occlusion affect image quality and model performance.
- **Data scarcity**: Limited annotated datasets for diverse crops and conditions hinder model training.
- **Computational constraints**: Real-time processing on edge devices requires efficient algorithms.
- **Integration complexity**: Combining vision systems with robotic platforms and decision support systems poses engineering challenges.

Despite these challenges, the survey concludes that computer vision holds great potential to revolutionize precision agriculture by enabling more efficient, sustainable, and productive farming practices. Future research should focus on developing robust, generalizable models and standardized evaluation protocols.

Who should read this

Pembaca Research Digest dan praktisi rumpun ini.

Opening member contentโ€ฆ