Penulis: Agil Evan; Moechammad Sarosa; Rosa Andrie Asmara; Mila Kusumawardani; Dimas Firmanda Al Riza; Yunia Mulyani Azis

Nama Jurnal: IEEE Xplore

Tahun: 2024

Volume: Tidak ada

Issue: Tidak ada

Halaman: Tidak ada

Deskripsi: By creating a mobile app framework that uses the YOLOv8 model to identify and count grape leaves. The goal is to improve grape (Vitis vinifera L.) growth monitoring by accurately counting leaves, which is critical for evaluating plant development. YOLOv8 is optimized for mobile applications using TensorFlow Lite, ensuring efficient processing. The methodology of this research includes dataset acquisition, annotation, and training of the YOLOv8 model with 100, 200, and 300 images. The model was then implemented on Android devices using TFLite to customize the performance of YOLOv8. The findings show that the model achieves up to 93% detection accuracy with the largest dataset (300 images) and mAP50-90 of 60%. Detection speed and accuracy are affected by dataset size, with larger datasets improving generalization but slightly slowing down inference time. The integration of YOLOv8 and TFLite into a mobile app provides an accessible and efficient tool for farmers to monitor crop growth by detecting and counting from leaves on grapes. This innovation has a significant impact on precision agriculture by enabling more accurate and timely decisionmaking, ultimately improving agricultural productivity and sustainability. The novelty of this research lies in the successful application of advanced deep learning models in a mobile framework, which offers practical solutions to real-world agricultural challenges.

Pendahuluan: In order to increase productivity and efficiency, modern agricultural production must find creative solutions to a number of problems. There are several obstacles to reaching ideal yields, including resource scarcity, pests, plant diseases, and climate change. Additionally, as attempts to fulfill the rising global food demand continue, there is an increasing need to promptly and correctly monitor agricultural conditions. In this situation, farmers require assistance in managing their fields more effectively and efficiently, which calls for the use of technology and techniques like precision agriculture [1], [2], [3]. a precision agriculture method that enhances crop yields by managing in-field variability via the use of information and communication technology. Precision agriculture enables farmers to make better educated decisions about irrigation, fertilization, and pest management by utilizing technology like sensors, GPS, and data analytics. Crop growth monitoring is a crucial part of precision farming as it may reveal information about how crops are developing [4], [5]. This monitoring aids in the early detection of issues as well as the efficient use of resources and increased productivity [6]. We use grape vines as an example of horticultural crops with significant economic value. Grapevine development and growth are significantly impacted by a variety of environmental elements and farming practices. Several variables contribute to this, such as the plant's daily morphological form and environmental circumstances. The quantity and quality of the leaves serve as a key sign of vine growth [7]. The process of photosynthesis, which controls the amount of energy available for plant growth and fruit production, is greatly influenced by leaves. Thus, keeping an eye on the quantity of grape leaves might provide important details about the plant's overall growth and even its general health [8], [9]. However, to make sure that plant development is not hampered, it has to be monitored quite regularly and thoroughly. It also needs free time because it still has to be manually watched and needs to be monitored on a regular basis. In order to lessen these issues, a plant growth forecast device based on automatically determining the number of leaves is required [10], [11]. Precision agricultural applications have demonstrated considerable potential for deep learning-based object recognition algorithms like YOLO (You Only Look Once) [12], [13]. Yolo is renowned for its high precision real-time item detecting capabilities. When considering vine growth, YOLO may be used to automatically identify and count the number of leaves, a crucial sign of the health and development of the plant. Leaf recognition and counting can be done more rapidly and precisely with the help of YOLOv8, the most recent version of the YOLO algorithm. YOLOv8 and TFLite have been integrated into smartphone applications, enabling farmers to use conveniently available devices to track the development of their crops in real time [14], [15], [16]. This study is critical for determining the effectiveness of the algorithm with the best dataset variation in recognizing and counting grape leaves. Plant growth monitoring and prediction will greatly benefit from an efficient and precise algorithm. According to earlier studies [17], [18] Faster RCNN can detect grape leaves with an accuracy of 90.11% and is also useful for identifying diseases in grape leaves. In order to construct a plant growth prediction engine based on automatic leaf count detection, the primary goal of this research is to obtain a grape leaf detection method. Leaf detection and quantity will be significantly impacted by the detection process's speed and accuracy. Growers may anticipate the health and growth of their grape plants by simply identifying and counting the number of leaves with the aid of an algorithm. As a result, there are less losses due to incorrect or delayed identification and increased production efficiency. These are the findings of this study, which may present new avenues for the advancement of agricultural science and technology in the future. It is anticipated that the combination of YOLOv8 and TFLite in a mobile application would offer a low-cost, high-efficiency method of tracking grapevine development, hence contributing to precision agriculture and sustainable agricultural outcomes.

Kata Kunci: Object Detection, Plant Growth, Agriculture, Android, YOLOv8

Total Kunjungan: 52 kali