Significant static & dynamic structures in natural scenes for visual recognition

明显的静电

• 批准号: 7526467
• 负责人: Jozsef Fiser
• 金额: $ 32.18万
• 依托单位: BRANDEIS UNIVERSITY
• 依托单位国家: 美国
• 财政年份: 2008
• 资助国家: 美国
• 起止时间: 2008-09-15 至 2012-08-31
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/7526467
• 关键词: Age related macular degeneration; Behavior; Blindness; Classification; Complex; Condition; Development; Dimensions; Disease; Elements; Environment; Eye; Eye Movements; Form Perception; Frequencies; Glaucoma; Goals; Grouping; Head; Higher Order Chromatin Structure; Human; Image; Image Analysis; Image Enhancement; Impairment; Individual; Intervention; Investigation; Masks; Measurement; Measures; Methodology; Methods; Motion; Outcome; Patients; Pattern; Perception; Peripheral; Population; Process; Psychophysiology; Public Health; Rehabilitation therapy; Research; Research Methodology; Residual state; Simulate; Speed; Stimulus; Structure; System; Techniques; Time; Title; Training; Vision; Visual; Visual Aid; Visual Perception; Visual impairment; Visual system structure; Work; abstracting; age related; base; cost; disorder of macula of retina; gaze; movie; novel; programs; tool; two-dimensional; visual process; visual processing

DESCRIPTION (provided by applicant): The overall goal of the proposed program of research is to identify the visual structures (complex features) and the dynamical tuning processes (adaptation) that humans utilize in perceiving their natural visual environment while they are actively engaged with it. The key element of our program is that we consider structure and relations between structures rather than pixel-based contrast patterns to be the fundamental representational primitives of vision and we develop a methodology to identify these primitives. We will define important structures in a statistical framework as significant higher order correlations among basic local measurements of image attributes (e.g. orientation, motion, contrast). We then combine rigorous psychophysical methods with new image manipulation techniques and with static and head-mounted eye tracking systems for a systematic analysis of the image features that support the identification and recognition of real and abstract objects. We will start with studies of perception of two-dimensional stationary images by normally-sighted people and expand our work to vision of freely inspected three-dimensional natural scenes and to people with visual impairment. The direct relevance of the present work to public health comes from the transitional component of the proposal. Our research methodology developed with visually-healthy subjects will be directly applied to two visually-impaired populations with the overall goal of informing the development of image enhancement techniques, effective low cost visual aids and visual training strategies to maximize visual function in patients with glaucoma and age-related macular disease. With the combination of basic and applied visual research, we hope to establish new and efficient approaches to rehabilitation of these visual diseases. Project Narrative The overall goal of the proposed program of research is to study how healthy humans and patients with visual deficiencies perceive their visual environment, what features they utilize in the during object and scene recognition. We will investigate the perception of still and moving images using psychophysical and eye-movement measures. Understanding these fundamental issues will allow us to develop specific image enhancement and rehabilitation techniques that can form the basis of effective visual aids and interventions for the visually impaired.

描述（申请人提供）：拟议研究计划的总体目标是确定视觉结构（复杂的特征）和动态调谐过程（适应），人类利用感知他们的自然视觉环境，而他们积极参与其中。我们计划的关键要素是，我们考虑结构和结构之间的关系，而不是像素-基于对比度模式是视觉的基本代表性原语，我们开发了一种方法来识别这些原语。我们将在统计框架中定义重要的结构，作为图像属性（例如方向，运动，对比度）的基本局部测量之间的显着高阶相关性。然后，我们结合联合收割机严格的心理物理方法与新的图像处理技术和静态和头戴式眼动跟踪系统的图像特征，支持识别和识别的真实的和抽象的对象进行系统的分析。我们将从正常视力的人对二维静止图像的感知研究开始，并将我们的工作扩展到自由观察的三维自然场景的视觉和视力障碍的人。目前的工作与公共卫生的直接相关性来自该提案的过渡部分。我们的研究方法与视觉健康的受试者开发将直接应用于两个视力受损的人群，告知图像增强技术的发展，有效的低成本的视觉辅助和视觉训练策略，以最大限度地提高青光眼和年龄相关性黄斑疾病患者的视觉功能的总体目标。我们希望通过视觉基础研究和应用研究的结合，为这些视觉疾病的康复建立新的、有效的方法。项目叙述拟议研究计划的总体目标是研究健康人和视觉缺陷患者如何感知他们的视觉环境，他们在物体和场景识别过程中利用了哪些功能。我们将使用心理物理学和眼动测量来研究静止和运动图像的感知。了解这些基本问题将使我们能够开发特定的图像增强和康复技术，这些技术可以为视障者提供有效的视觉辅助和干预措施。