SBIR Phase II: Inexpensive Automatic Classification And Counting Of Insects To Enable Precision Agriculture

这项小企业创新研究（SBIR）第二阶段项目的更广泛影响是在提高作物产量的同时减少农药的使用。昆虫每年破坏或毁坏约1500亿美元的农作物。提高昆虫监测的准确性和及时性将有助于更有效地管理害虫，使昆虫干预在空间和时间上有针对性；例如，拟议的系统可以建议在一天的最佳时间在选定的“热点”喷洒一种温和（和便宜）的农药，而不是在整个田地上广泛喷洒刺激性农药。通过减少污染和潜在地减轻蜂群衰竭失调症，减少杀虫剂的数量具有进一步的积极效益。本项目创建的硬件、算法、表示和数据模型可广泛应用于蚊虫监测，对控制人类和牲畜的虫媒疾病具有重要意义。拟议的项目将推进最先进的飞虫分类，以改善精确农业的昆虫监测为目标。该研究将推进在野外遇到的各种条件（温度、压力、湿度）的算法和表示的使用，因为这些条件会影响空气密度，进而影响昆虫的飞行。目前的模型使用密度和升力的标准模型，将昆虫视为理想的空气动力学对象，忽略了环境对昆虫生理的影响。该项目将使用机器学习来提高模型的准确性和精度。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

The broader impact of this Small Business Innovation Research (SBIR) Phase II project is in improving crop yields while reducing pesticide use. Insects damage or destroy about $150 B of crops each year. Improving the accuracy and timeliness of insect surveillance will allow more effective pest management, allowing the insect interventions to be targeted in space and time; for example, rather than broadly spraying harsh pesticides across an entire field, the proposed system could suggest spraying of a milder (and cheaper) pesticide in select “hot spots” at the optimal time of day. Reducing the volume of pesticides has further positive benefits by reducing pollution and potentially mitigating colony collapse disorder. The hardware, algorithms, representations, and data models created in this project can be applied broadly to mosquito surveillance, with implications for control of insect-vectored diseases of both humans and livestock. The proposed project will advance the state-of-the-art in flying insect classification, with the goal of improving insect surveillance for precision agriculture. The study will advance the use of algorithms and representations for a wide range of conditions (temperature, pressure, humidity) encountered in the field, as these conditions affect air density, which in turn impacts insect flight. Current models use standard models for density and lift, treating insects as idealized aerodynamic objects and ignoring effects of the environment on insect physiology. This project will use machine learning to improve model accuracy and precision.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.