PFI-TT: Interactive Software for Hyperspectral Image Analysis

PFI-TT：用于高光谱图像分析的交互式软件

• 批准号: 1827656
• 负责人: Mikhail Berezin
• 金额: $ 20万
• 依托单位: Washington University
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2018
• 资助国家: 美国
• 起止时间: 2018-08-01 至 2022-07-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1827656&HistoricalAwards=false
• 关键词: PFI; TT; Interactive; Software; Hyperspectral

The broader impact/commercial potential of this PFI project is to advance, facilitate and streamline the analysis of hyperspectral imaging data. In contrast to two-dimensional images produced by conventional color cameras usually with three-color channels, hyperspectral techniques generate three-dimensional datasets known as datacubes with both spatial and spectral dimensions with hundreds of color channels. Hyperspectral methods lead to a vastly improved ability to classify and differentiate objects based on their spectral features enabling small, otherwise unnoticeable, features to be amplified. The societal impact of hyperspectral imaging is rapidly expanding and includes the development of advanced medical diagnostic tools, systems for precision agriculture, methods to ensure food quality, discovery of new minerals, and the evolution of national defense projects. However, despite the progress in analysis method development and computational speed, efficient computational approaches suitable for rapid analysis of datacubes are still lacking. Three-dimensionality of data and large file size presents a significant challenge for isolation of useful data, and consequently individual investigators developing their own tools. The development of an advanced and commercially available software package for hyperspectral data analysis will facilitate this work and enable users to search for new optical signatures across multiple applications. The proposed project sets a new software architecture that is capable of analyzing complex and computationally challenging datasets recorded by hyperspectral imaging systems. The proposed computational platform for data analysis of hyperspectral data will provide a unique and powerful tool to process hyperspectral data for a variety of applications: from medicine and remote vision to forensics and plants science. The research objective is to integrate current and new analysis methods into an interactive, fast, and intuitive hyperspectral imaging software that will offer an unprecedented level of flexibility in hyperspectral data processing. The proposed software program will be designed to be operated without any prior programming skills allowing interactive sessions of raw and processed data to be easily shared, explored and evaluated by both peers and newcomers. The major feature of the software is an open architecture that provides a common framework for the implementation of the additional methods from the user community. The commercially available platform will have structural flexibility to allow users to integrate their novel computational methods with relatively little effort. With this feature, the software is expected to be a central part of hyperspectral technology and will be the preferred software of the majority of users.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.

该PFI项目更广泛的影响/商业潜力是推进、促进和简化高光谱成像数据的分析。与传统彩色相机通常使用三色通道生成的二维图像相比，高光谱技术生成的三维数据集称为数据立方体，具有数百个颜色通道的空间和光谱维度。高光谱方法极大地提高了根据光谱特征对物体进行分类和区分的能力，使微小的、不明显的特征得以放大。高光谱成像的社会影响正在迅速扩大，包括先进医疗诊断工具的发展、精准农业系统、确保食品质量的方法、新矿物的发现以及国防项目的发展。然而，尽管分析方法的发展和计算速度都有了很大的进步，但仍然缺乏适合于快速分析数据立方体的高效计算方法。数据的三维性和大文件大小对隔离有用数据提出了重大挑战，因此个别调查人员开发了自己的工具。开发用于高光谱数据分析的先进商用软件包将促进这项工作，并使用户能够在多种应用中搜索新的光学特征。拟议的项目设置了一个新的软件架构，能够分析高光谱成像系统记录的复杂和计算挑战性的数据集。提出的用于高光谱数据分析的计算平台将提供一个独特而强大的工具来处理各种应用的高光谱数据：从医学和远程视觉到法医学和植物科学。研究目标是将当前和新的分析方法集成到一个交互式、快速和直观的高光谱成像软件中，该软件将为高光谱数据处理提供前所未有的灵活性。拟议的软件程序将被设计成在没有任何先前编程技能的情况下操作，允许原始和处理数据的交互会话很容易被同行和新手共享、探索和评估。该软件的主要特性是一个开放的体系结构，它为实现来自用户社区的其他方法提供了一个公共框架。商业上可用的平台将具有结构灵活性，允许用户以相对较少的努力集成他们的新计算方法。有了这个特性，该软件有望成为高光谱技术的核心部分，并将成为大多数用户的首选软件。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。

项目成果

IDCube Lite: Free Interactive Discovery Cube software for multi and hyperspectral applications

IDCube Lite：适用于多光谱和高光谱应用的免费交互式 Discovery Cube 软件

• DOI: 10.1255/jsi.2021.a1
• 发表时间: 2021
• 期刊: Journal of Spectral Imaging
• 作者: Mishra, Deependra;Hurbon, Helena;Wang, John;Wang, Steven;Du, Tommy;Wu, Qian;Kim, David;Basir, Shiva;Cao, Qian;Zhang, Hairong
• 通讯作者: Zhang, Hairong

HSKL: A Machine Learning Framework for Hyperspectral Image Analysis

• DOI: 10.1109/whispers52202.2021.9483968
• 发表时间: 2021-03
• 期刊: 2021 11th Workshop on Hyperspectral Imaging and Signal Processing: Evolution in Remote Sensing (WHISPERS)
• 作者: Qian Cao;Deependra Mishra;John Wang;Steven T. Wang;Helena Hurbon;M. Berezin

Idcube Lite – A Free Interactive Discovery Cube Software for Multi And Hyperspectral Applications

Idcube Lite — 适用于多光谱和高光谱应用的免费交互式 Discovery Cube 软件

• 发表时间: 2021
• 期刊: Workshop on Hyperspectral Image and Signal Processing Evolution in Remote Sensing
• 作者: Mishra, D;Hurbon, H;Wang, J;Wang, T;Kim, D;Cao, Q;Basir, S;Zhang, H
• 通讯作者: Zhang, H

Selection of Hyperspectral Endmember Extraction Algorithm for Tumor Delineation in Animal Models

用于动物模型肿瘤勾画的高光谱端元提取算法的选择

• 发表时间: 2021
• 期刊: paper OF2E.2
• 作者: Mishra, D;Wang, J;Wang, ST;Cao, Q;Hurbon, H;Akers, W;Berezin, MY
• 通讯作者: Berezin, MY