Farmers’ Toolkit: Deep Learning in Weed Detection and Precision Crop & Fertilizer Recommendations
Agriculture is widely recognized as a significant and indispensable occupation on a global scale. The current imperative is to optimize agricultural practices and progressively transition towards smart agriculture. The Internet of Things (IoT) technology has dramatically enhanced people’s daily live...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
EDP Sciences
2024-01-01
|
| Series: | BIO Web of Conferences |
| Online Access: | https://www.bio-conferences.org/articles/bioconf/pdf/2024/01/bioconf_msnbas2024_05012.pdf |
| _version_ | 1797353065973547008 |
|---|---|
| author | Dash Sushree Sasmita Kumar Pawan |
| author_facet | Dash Sushree Sasmita Kumar Pawan |
| author_sort | Dash Sushree Sasmita |
| collection | DOAJ |
| description | Agriculture is widely recognized as a significant and indispensable occupation on a global scale. The current imperative is to optimize agricultural practices and progressively transition towards smart agriculture. The Internet of Things (IoT) technology has dramatically enhanced people’s daily lives via diverse applications across several domains. Previous studies have yet to effectively incorporate Artificial Intelligence (AI) with sensor technology to provide comprehensive guidance to agricultural practitioners, hindering their ability to achieve good outcomes. This research offers Farmers’ Toolkit with four layers: sensor, network, service, and application. This toolkit aims to facilitate the implementation of a smart farming system while effectively managing energy resources. With a specific emphasis on the application layer, the toolkit uses a deep learning methodology to construct a fertilizer recommendation system that aligns with the expert’s perspective. This study utilizes IoT devices and Wireless Sensor Network (WSN) methods to enhance the efficiency and speed of recommending appropriate crops to farmers. The recommendation process considers several criteria: temperature, yearly precipitation, land area, prior crop history, and available resources. The identification of undesirable vegetation on agricultural fields, namely the detection of weeds, is carried out using drone technology equipped with frame-capturing capabilities and advanced deep-learning algorithms. The findings demonstrate an accuracy rate of 94%, precision rate of 92%, recall rate of 96%, and F1 score of 94%. The toolkit for farmers alleviates physical labor and time expended on various agricultural tasks while enhancing overall land productivity, mitigating potential crop failures in specific soil conditions, and minimizing crop damage inflicted by weeds. |
| first_indexed | 2024-03-08T13:24:50Z |
| format | Article |
| id | doaj.art-ae94e677d9d34d4e9d5c9610aaa04d02 |
| institution | Directory Open Access Journal |
| issn | 2117-4458 |
| language | English |
| last_indexed | 2024-03-08T13:24:50Z |
| publishDate | 2024-01-01 |
| publisher | EDP Sciences |
| record_format | Article |
| series | BIO Web of Conferences |
| spelling | doaj.art-ae94e677d9d34d4e9d5c9610aaa04d022024-01-17T14:59:09ZengEDP SciencesBIO Web of Conferences2117-44582024-01-01820501210.1051/bioconf/20248205012bioconf_msnbas2024_05012Farmers’ Toolkit: Deep Learning in Weed Detection and Precision Crop & Fertilizer RecommendationsDash Sushree Sasmita0Kumar Pawan1Faculty of CS & IT, Kalinga UniversityFaculty of CS & IT, Kalinga UniversityAgriculture is widely recognized as a significant and indispensable occupation on a global scale. The current imperative is to optimize agricultural practices and progressively transition towards smart agriculture. The Internet of Things (IoT) technology has dramatically enhanced people’s daily lives via diverse applications across several domains. Previous studies have yet to effectively incorporate Artificial Intelligence (AI) with sensor technology to provide comprehensive guidance to agricultural practitioners, hindering their ability to achieve good outcomes. This research offers Farmers’ Toolkit with four layers: sensor, network, service, and application. This toolkit aims to facilitate the implementation of a smart farming system while effectively managing energy resources. With a specific emphasis on the application layer, the toolkit uses a deep learning methodology to construct a fertilizer recommendation system that aligns with the expert’s perspective. This study utilizes IoT devices and Wireless Sensor Network (WSN) methods to enhance the efficiency and speed of recommending appropriate crops to farmers. The recommendation process considers several criteria: temperature, yearly precipitation, land area, prior crop history, and available resources. The identification of undesirable vegetation on agricultural fields, namely the detection of weeds, is carried out using drone technology equipped with frame-capturing capabilities and advanced deep-learning algorithms. The findings demonstrate an accuracy rate of 94%, precision rate of 92%, recall rate of 96%, and F1 score of 94%. The toolkit for farmers alleviates physical labor and time expended on various agricultural tasks while enhancing overall land productivity, mitigating potential crop failures in specific soil conditions, and minimizing crop damage inflicted by weeds.https://www.bio-conferences.org/articles/bioconf/pdf/2024/01/bioconf_msnbas2024_05012.pdf |
| spellingShingle | Dash Sushree Sasmita Kumar Pawan Farmers’ Toolkit: Deep Learning in Weed Detection and Precision Crop & Fertilizer Recommendations BIO Web of Conferences |
| title | Farmers’ Toolkit: Deep Learning in Weed Detection and Precision Crop & Fertilizer Recommendations |
| title_full | Farmers’ Toolkit: Deep Learning in Weed Detection and Precision Crop & Fertilizer Recommendations |
| title_fullStr | Farmers’ Toolkit: Deep Learning in Weed Detection and Precision Crop & Fertilizer Recommendations |
| title_full_unstemmed | Farmers’ Toolkit: Deep Learning in Weed Detection and Precision Crop & Fertilizer Recommendations |
| title_short | Farmers’ Toolkit: Deep Learning in Weed Detection and Precision Crop & Fertilizer Recommendations |
| title_sort | farmers toolkit deep learning in weed detection and precision crop fertilizer recommendations |
| url | https://www.bio-conferences.org/articles/bioconf/pdf/2024/01/bioconf_msnbas2024_05012.pdf |
| work_keys_str_mv | AT dashsushreesasmita farmerstoolkitdeeplearninginweeddetectionandprecisioncropfertilizerrecommendations AT kumarpawan farmerstoolkitdeeplearninginweeddetectionandprecisioncropfertilizerrecommendations |