Digital image processing is one of the potential technologies used in precision agriculture to gather information, such as seed emergence, plant health, and phenology from the digital images. Despite its potential, the rate of adoption is slow due to limited accessibility, unsuitability to specific issues, unaffordability, and high technical knowledge requirement from the clientele. Therefore, the development of open source image processing applications that are task-specific, easy-to-use, requiring fewer inputs, and rich with features will be beneficial to the users/farmers for adoption. The Fiji software, an open source free image processing ImageJ platform, was used in this application development study. A collection of four different agricultural field applications were selected to address the existing issues and develop image processing tools by applying novel approaches and simple mathematical principles. First, an automated application, using a digital image and “pixel-march” method, performed multiple radial measurements of sunflower floral components. At least 32 measurements for ray florets and eight for the disc were required statistically for accurate dimensions. Second, the color calibration of digital images addressed the light intensity variations of images using standard calibration chart and derived color calibration matrix from selected color patches. Calibration using just three-color patches: red, green, and blue was sufficient to obtain images of uniform intensity. Third, plant stand count and their spatial distribution from UAS images were determined with an accuracy of ≈96 %, through pixel-profile identification method and plant cluster segmentation. Fourth, the soybean phenological stages from the PhenoCam time-lapse imagery were analyzed and they matched with the manual visual observation. The green leaf index produced the minimum variations from its smoothed curve. The time of image capture and PhenoCam distances had significant effects on the vegetation indices analyzed. A simplified approach using kymograph was developed, which was quick and efficient for phenological observations. Based on the study, these tools can be equally applied to other scenarios, or new user-coded, user-friendly, image processing tools can be developed to address specific requirements. In conclusion, these successful results demonstrated the suitability and possibility of task-specific, open source, digital image processing tools development for agricultural field applications. / United States. Agricultural Research Service / National Institute of Food and Agriculture (U.S.)

Agriculture faces many challenges related to the increasing food demands of a growing global population and the sustainable use of resources in a changing environment. To address them, we need reliable information sources, like exploiting hyperspectral satellite, airborne, and ground-based remote sensing data to observe phenological traits through a crops growth cycle and gather information to precisely diagnose when, why, and where a crop is suffering negative impacts. By combining hyperspectral capabilities with unmanned aerial vehicles (UAVs), there is an increased capacity for providing time-critical monitoring and new insights into patterns of crop development. However, considerable effort is required to effectively utilize UAV-integrated hyperspectral systems in crop-modeling and crop-breeding tasks. Here, a UAV-based hyperspectral solution for mapping crop physiological parameters was explored within a machine learning framework. To do this, a range of complementary measurements were collected from a field-based phenotyping experiment, based on a diversity panel of wild tomato (Solanum pimpinellifolium) that were grown under fresh and saline conditions. From the UAV data, positionally accurate reflectance retrievals were produced using a computationally robust automated georectification and mosaicking methodology. The resulting multitemporal UAV data were then employed to retrieve leaf-chlorophyll (Chl) dynamics via a machine learning framework. Several approaches were evaluated to identify the best-performing regression supervised methods. An investigation of two learning strategies (i.e., sequential and retraining) and the value of using spectral bands and vegetation indices (VIs) as prediction features was also performed. Finally, the utility of UAVbased hyperspectral phenotyping was demonstrated by detecting the effects of salt-stress on the different tomato accessions by estimating the salt-induced senescence index from the retrieved Chl dynamics, facilitating the identification of salt-tolerant candidates for future investigations. This research illustrates the potential of UAV-based hyperspectral imaging for plant phenotyping and precision agriculture. In particular, a) developing systematic imaging calibration and pre-processing workflows; b) exploring machine learning-driven tools for retrieving plant phenological dynamics; c) establishing a plant stress detection approach from hyperspectral-derived metrics; and d) providing new insights into using computer vision, big-data analytics, and modeling strategies to deal effectively with the complexity of the UAV-based hyperspectral data in mapping plant physiological indicators.

No description available.

The collective body of U.S. legislation, colloquially known as the Farm Bill, authorizes a suite of practices and programs under its Conservation Title. This includes the Conservation Reserve Program (CRP), which incentivizes agricultural producers to remove arable land from production to enhance soil retention, improve water quality, and restore wildlife habitat. Conservation Practice 33: Habitat Buffers for Upland Birds (CP-33) was the first CRP practice to target wildlife habitat restoration. CP-33 incentivizes producers to reestablish native herbaceous vegetation along crop field margins. Producers are often concerned with the economic opportunity costs of CP-33 enrollment. I used yield data derived from 44 agricultural fields in the Mississippi Alluvial Valley, USA to compare the environmental and economic opportunities associated with CP-33 establishment. I used yield data to develop a revenue distribution function to illustrate CP-33 revenue as commodity prices fluctuate. I found that as commodity prices increase, CP-33 implementation becomes less profitable.

Intense agriculture is detrimental to the environment and leads to nutrient runoff, decreased water quality, soil erosion, greenhouse gas emissions, and decreased wildlife habitat. In addition to negative environmental impacts, intense agriculture increases the financial strain on farmers. Farmers also face inconsistent yields from environmentally vulnerable lands. Due to these financial constraints and inconsistent yields, conservation goals are not always economically attractive to farmers and agricultural producers. One possible solution to this issue is the use of precision agriculture (PA) to inform strategic conservation efforts. We used PA technology to identify low-revenue field areas in the Mississippi Delta and Black Prairie regions. We created spatially explicit revenue maps and overlaid it with the Biologist Ranking Index (BRI) to illustrate where economic and conservation opportunities overlap. Our results indicate that, on average, upwards of 20.1% of the Black Prairie and 18.0% of the Mississippi Delta generate less revenue than conservation enrollment.

Indiana University-Purdue University Indianapolis (IUPUI) / Detecting crop rows from video frames in real time is a fundamental challenge in the field of precision agriculture. Deep learning based semantic segmentation method, namely U-net, although successful in many tasks related to precision agriculture, performs poorly for solving this task. The reasons include paucity of large scale labeled datasets in this domain, diversity in crops, and the diversity of appearance of the same crops at various stages of their growth. In this work, we discuss the development of a practical real-life crop row detection system in collaboration with an agricultural sprayer company. Our proposed method takes the output of semantic segmentation using U-net, and then apply a clustering based probabilistic temporal calibration which can adapt to different fields and crops without the need for retraining the network. Experimental results validate that our method can be used for both refining the results of the U-net to reduce errors and also for frame interpolation of the input video stream. Upon the availability of more labeled data, we switched our approach from a semi-supervised model to a fully supervised end-to-end crop row detection model using a Feature Pyramid Network or FPN. Central to the FPN is a pyramid pooling module that extracts features from the input image at multiple resolutions. This results in the network’s ability to use both local and global features in classifying pixels to be crop rows. After training the FPN on the labeled dataset, our method obtained a mean IoU or Jaccard Index score of over 70% as reported on the test set. We trained our method on only a subset of the corn dataset and tested its performance on multiple variations of weed pressure and crop growth stages to verify that the performance does translate over the variations and is consistent across the entire dataset.

Internet of Things (IoT) has drawn much attention in recent years. With IoT, physical world entities get connected through internet. IoT is used currently in various applications, such as environmental monitoring, control systems, farming, home automation, security and surveillance systems etc. The aim of this research is to design a low-cost, energy-efficient, reliable and scalable embedded application for the smart greenhouse monitoring system. The IoT based system designed in this thesis uses various sensors to measure the air and soil quality parameters in the greenhouse, and monitor real-time data online using web-server and mobile phone based applications. A ZigBee based wireless sensor network is implemented to transport various sensory data to the gateway. Among other contributions, the designed system develops a new routing algorithm by introducing a confirmed delivery of packets and re-routing features. We also introduced an efficient cost metric for making routing decisions within WSN using hops count, and simple bi-directional link quality estimator using PRR and current battery voltage of neighbor nodes. We also verified the stability of the system by conducting various performance tests. The system is equipped with data analytic functions for the online examination of the data. The designed system adopts event-based triggering and data aggregation methods to reduce the number of transmissions, and develops a new algorithm for such purpose. The web-server and mobile applications have user interface to display the output of the data analytic services, warning, control operations and give access to query data of the user's interest.

The Internet of Things (IoT) is an umbrella term for smart things connected to the Internet.Precision agriculture is a related concept where connected sensors can be used to facilitate, e.g. more effective farming. At the same time, Bluetooth has been making advancements into IoT with the release of Bluetooth Low Energy (BLE) or Bluetooth smart as it is also known by. This thesis describes the development of a Bluetooth Low Energy moisture- and temperature sensor intended for use in an agricultural wireless sensor network system. The sensor was evaluated based on its effectiveness in agricultural environments and conditions such as weather, elevation and in different crop fields. Bluetooth Low Energy was chosen as the technology for communication by the supervising company due to its inherent support for mobile phone accessibility.Field tests showed that the sensor nodes were largely affected by greenery positioned betweentransmitter and receiver, meaning that these would preferably be placed above growing crops foreffective communication. With ideal placement of the sensor and receiving unit, the signal wouldreach up to 100 m, meaning that a receiving unit would cover a circle area with radius 100 m.Due to Bluetooth being largely integrated in mobile devices it would mean that sensor data couldeasily be made accessible with a mobile app, rather than acquiring data from an online web server.

Obtaining accurate and precise position information is critical in precision and autonomous agriculture. Systems accurate to the centimeter-level are available, but may be prohibitively expensive for relatively small farms and tasks that involve multiple vehicles. Optical resection is proposed as a potentially more cost-effective and scalable positioning system for such cases. The proposed system involves the placement of optical beacons at known locations throughout the environment and the use of cameras on the vehicle to detect the apparent angles between beacons. The position of the vehicle can be calculated with resection when three or four beacons are identified. In addition, the system provides precise orientation information, so a separate inertial measurement unit is not required. The system is seen as potentially cost-effective by taking advantage of the precision and low cost of digital image sensors. Whereas the components in other positioning systems tend to be more specialized, the widespread consumer demand for inexpensive and high quality cameras has allowed for billions of dollars of research and development to be spread across billions of image sensors. / Master of Science

Effective weed control is crucial, especially for herbicide-resistant species. Machine vision technology, through weed detection and localization, can facilitate precise, species-specific treatments. Despite the challenges posed by unstructured field conditions and weed variability, deep learning (DL) algorithms show promise. This study evaluated thirteen DL-based weed detection models, including YOLOv5, RetinaNet, EfficientDet, Fast RCNN, and Faster RCNN, using pre-trained object detectors. RetinaNet (R101-FPN) achieved the highest accuracy with a mean average precision (mAP@0.50) of 79.98%, though it had longer inference times. YOLOv5n, with the fastest inference (17 ms on Google Colab) and only 1.8 million parameters, achieved a comparable 76.58% mAP@0.50, making it suitable for real-time use in resource-limited devices. A prototype using YOLOv5 was tested on two datasets, showing good real-time accuracy on In-season data and comparable results on Cross-season data, despite some accuracy challenges due to dataset distribution shifts.