Optical techniques for quality determination of cereal grains

导语：在当今世界，消费者越来越关注他们食品的安全和质量。这给生产行业带来了提高质量监控标准的压力。质量检验流程的自动化是该行业能够提供一致和高效结果的唯一途径。然而，在提供廉价和实时解决方案以满足这些需求方面存在技术差距。它假设数学的基本理论和现代计算技术可以结合在一起，并用于开发食品质量综合评价仪器。科学方法：我的研究的长期目标是发展光谱信号与食品材料质量方面相关的基础的高级知识；改进硬件和软件技术以增强这些信号；并利用这些信息开发仪器来优化食品、营养食品、乳制品等行业的质量监测、分级和加工操作。拟议的研究集中于开发和优化算法和数据分析技术，以获取、增强、转换和分析小麦籽粒的高光谱近红外图像。开发的方法将在加拿大西部的小麦上进行测试，以确定早期发病的镰刀菌，这种霉菌会对人和动物的健康造成危害，并给全球谷物行业造成数百万美元的经济损失。光谱数据将使用小波、多变量图像分析和遗传算法等技术进行分析。预期意义：通过本研究开发的知识库将为整个社会带来健康效益，为加拿大粮食产业带来经济效益。虽然在这项研究中，所开发的技术仅在食品工程领域进行评估，但它们将在其他各种学科中发挥作用，如制造业、生物医学、纳米技术、遥感等。

INTRODUCTION: In today's world, consumers are becoming increasingly concerned about the safety and quality of their food. This has put a pressure on the production industry to heighten quality-monitoring standards. Automation of quality inspection processes is the only way the industry can deliver consistent and efficient results. However, there is gap in technology to provide inexpensive and real-time solutions to meet these demands. It is hypothesized that the fundamental theories of mathematics and modern computational techniques can be integrated together and be used for developing instrumentation for comprehensive assessment of food quality. SCIENTIFIC APPROACH: The long-term goal of my research is to develop advanced knowledge of the basis by which spectral signals are related to the quality aspects of food materials; improve hardware and software techniques to enhance these signals; and use this information to develop instrumentation to optimize the quality monitoring, grading, and processing operations in industries such as food, nutraceutical, dairy, etc.The proposed research is focussed on developing and optimizing algorithms and data analyses techniques to acquire, enhance, transform, and analyze hyperspectral near-infrared images of wheat kernels. The developed methods will be tested on western Canadian wheat to identify early onset of Fusarium, a mold that causes health hazards in humans and animals and causes millions of dollars of economic losses to the grain industry worldwide. Spectral data will be analyzed using techniques such as wavelets, multivariate-image analysis, and genetic algorithm. EXPECTED SIGNIFICANCE: The knowledge base developed through this research will provide health benefits to the whole society and economic benefits to the Canadian grain industry. Although the developed techniques are being evaluated only in the food engineering area in this study, they will be instrumental in a variety of other disciplines e.g. manufacturing, biomedicine, nanotechnology, remote sensing, etc.