Vision-based tracking for sports performance analysis

With the world of professional sports shifting towards employing better sport analytics, the demand for vision-based performance analysis is growing increasingly in recent years. In addition, the nature of many sports does not allow the use of any kind of sensors or other wearable markers attached to players for monitoring their performances during competitions. This provides a potential application of systematic observations such as tracking information of the players to help coaches to develop their visual skills and perceptual awareness needed to make decisions about team strategy or training plans. My PhD project is part of a bigger ongoing project between sport scientists and computer scientists involving also industry partners and sports organisations. The overall idea is to investigate the contribution technology can make to the analysis of sports performance on the example of team sports such as rugby, football or hockey. A particular focus is on vision-based tracking, so that information about the location and dynamics of the players can be gained without any additional sensors on the players. To start with, prior approaches on visual tracking are extensively reviewed and analysed. In this thesis, methods to deal with the difficulties in visual tracking to handle the target appearance changes caused by intrinsic (e.g. pose variation) and extrinsic factors, such as occlusion, are proposed. This analysis highlights the importance of the proposed visual tracking algorithms, which reflect these challenges and suggest robust and accurate frameworks to estimate the target state in a complex tracking scenario such as a sports scene, thereby facilitating the tracking process. Next, a framework for continuously tracking multiple targets is proposed. Compared to single target tracking, multi-target tracking such as tracking the players on a sports field, poses additional difficulties, namely data association, which needs to be addressed. Here, the aim is to locate all targets of interest, inferring their trajectories and deciding which observation corresponds to which target trajectory is. In this thesis, an efficient framework is proposed to handle this particular problem, especially in sport scenes, where the players of the same team tend to look similar and exhibit complex interactions and unpredictable movements resulting in matching ambiguity between the players. The presented approach is also evaluated on different sports datasets and shows promising results. Finally, information from the proposed tracking system is utilised as the basic input for further higher level performance analysis such as tactics and team formations, which can help coaches to design a better training plan. Due to the continuous nature of many team sports (e.g. soccer, hockey), it is not straightforward to infer the high-level team behaviours, such as players’ interaction. The proposed framework relies on two distinct levels of performance analysis: low-level performance analysis, such as identifying players positions on the play field, as well as a high-level analysis, where the aim is to estimate the density of player locations or detecting their possible interaction group. The related experiments show the proposed approach can effectively explore this high-level information, which has many potential applications.

Obstacle-free range determination for rail track maintenance vehicles

Maintenance trains travel in convoy. In Australia, only the first train of the convoy pays attention to the track sig- nalization (the other convoy vehicles simply follow the preceding vehicle). Because of human errors, collisions can happen between the maintenance vehicles. Although an anti-collision system based on a laser distance meter is already in operation, the existing system has a limited range due to the curvature of the tracks. In this paper, we introduce an anti-collision system based on vision. The two main ideas are, (1) to warp the camera image into an image where the rails are parallel through a projective transform, and (2) to track the two rail curves simultaneously by evaluating small parallel segments. The performance of the system is demonstrated on an image dataset.

The visual and functional impacts of astigmatism and its clinical management

Purpose: To provide a comprehensive overview of research examining the impact of astigmatism on clinical and functional measures of vision, the short and longer term adaptations to astigmatism that occur in the visual system, and the currently available clinical options for the management of patients with astigmatism. Recent findings: The presence of astigmatism can lead to substantial reductions in visual performance in a variety of clinical vision measures and functional visual tasks. Recent evidence demonstrates that astigmatic blur results in short-term adaptations in the visual system that appear to reduce the perceived impact of astigmatism on vision. In the longer term, uncorrected astigmatism in childhood can also significantly impact on visual development, resulting in amblyopia. Astigmatism is also associated with the development of spherical refractive errors. Although the clinical correction of small magnitudes of astigmatism is relatively straightforward, the precise, reliable correction of astigmatism (particularly high astigmatism) can be challenging. A wide variety of refractive corrections are now available for the patient with astigmatism, including spectacle, contact lens and surgical options. Conclusion: Astigmatism is one of the most common refractive errors managed in clinical ophthalmic practice. The significant visual and functional impacts of astigmatism emphasise the importance of its reliable clinical management. With continued improvements in ocular measurement techniques and developments in a range of different refractive correction technologies, the future promises the potential for more precise and comprehensive correction options for astigmatic patients.

Retinal anatomy of the New Zealand kiwi: Structural traits consistent with their nocturnal behavior

Kiwi (Apteryx spp.) have a visual system unlike that of other nocturnal birds, and have specializations to their auditory, olfactory and tactile systems. Eye size, binocular visual fields and visual brain centers in kiwi are proportionally the smallest yet recorded among birds. Given the many unique features of the kiwi visual system, we examined the laminar organization of the kiwi retina to determine if they evolved increased light sensitivity with a shift to a nocturnal niche or if they retained features of their diurnal ancestor. The laminar organization of the kiwi retina was consistent with an ability to detect low light levels similar to that of other nocturnal species. In particular, the retina appeared to have a high proportion of rod photoreceptors compared to diurnal species, as evidenced by a thick outer nuclear layer, and also numerous thin photoreceptor segments intercalated among the conical shaped cone photoreceptor inner segments. Therefore, the retinal structure of kiwi was consistent with increased light sensitivity, although other features of the visual system, such as eye size, suggest a reduced reliance on vision. The unique combination of a nocturnal retina and smaller than expected eye size, binocular visual fields and brain regions make the kiwi visual system unlike that of any bird examined to date. Whether these features of their visual system are an evolutionary design that meets their specific visual needs or are a remnant of a kiwi ancestor that relied more heavily on vision is yet to be determined.

Inarinsaamen käsitemetaforat

This study discusses the conceptual metaphors of Inari Saami, an endangered, indigenous, Finno-Ugrian language spoken in northern Finland. The research focuses on systematical mappings between source and target domains in conventional Inari Saami metaphors and metonymies. The research material consists of the Inarinsaamen idiomisanakirja [Inari Saami idiom dictionary] which has been compiled by the author in collaboration with an Inari Saami co-author; the Inarilappisches Wörterbuch; Inarinsaamelaista kansantietoutta [Inari Saami folk knowledge]; and Aanaarkiela čájttuzeh [Inari Saami sample texts]. The metaphors and metonymies found in these literary sources are divided into categories on the basis of the target domains and according to the classic model of Lakoff ja Johnson (1980). This method reveals the systematical recurrence of source domains inside each category and thus discovers the systematical patterns of metaphoric mapping, the conceptual metaphors . As a result 44 conceptual metaphors and 16 conceptual metonymies are presented through approximately 500 glossed examples. These findings are discussed against the background of what is known about the cognitive and neural processing of metaphors on the one hand, and what is known about Inari Saami culture on the other. This theoretical framework highlights culture as the underlying force behind conceptual metaphors. The recurring metonymies seem to follow a culturally salient indexicality. For example, the Inari Saami conceptual metonymy TIME IS NATURE reflects the seasonal changes in the year s cycle, which was the salient index of time in traditional Inari Saami culture. The recurring metaphors, for their part, follow a culturally salient iconicity. The conceptual metaphor PRIDE IS ANTLERS is based on an iconicity which is experienced and interpreted by the Inari Saami. A proud person is associated with a reindeer who shows off his impressive antlers. The conceptual metaphor/metonymy seems to be a reflection of culture rather than a cognitive means of understanding an abstract domain in terms of a concrete domain, as hypothesized by certain theoreticians. Repeating this study with other languages may lead to the possibility of typologizing the metaphorical systems of the world s languages and understanding the diversity of metaphor systems in the endangered languages of the world.

Reconstruction of the ocular surface using biomaterial templates

This chapter discusses the effect on vision of a large group of pathological conditions, known as ocular surface disorders (OSDs), and presents the therapeutic strategies to reconstruct the abnormal ocular surface. If left untreated, most of the OSDs will lead to partial or total loss of eyesight, especially when limbal stem cell deficiency is involved. An overview of various treatment strategies is presented, with the emphasis on the development of the ex vivo expansion of corneal limbal epithelial cells (presumed to be progenitor or stem cells) and the creation of transplantable epithelial constructs. The use of naturally derived biomaterials (collagen, fibrin, amnion, etc.) or synthetic polymers (polylactides, thermoresponsive polymers, etc.) as substrata in these constructs is critically analyzed. Emphasis is placed on the templates from silk proteins, which are being developed by the authors.

Object Tracking Using Motion Direction Detection

We present an algorithm for tracking objects in a video sequence, based on a novel approach for motion detection. We do not estimate the velocity �eld. In-stead we detect only the direction of motion at edge points and thus isolate sets of points which are moving coherently. We use a Hausdor� distance based matching algorithm to match point sets in local neighborhood and thus track objects in a video sequence. We show through some examples the e�ectiveness of the algo- rithm.

The role of sense organs in the feeding behaviour of Chinese perch

Experiments were conducted to identify the rules of the individual sense organs in the feeding behaviour of Chinese perch Siniperca chuatsi by determining the consumption of natural food after selective removal or blocking of eyes, lateral lines and olfactory organs, and also by observing the behavioural response to visual, mechanical and chemical stimulation by artificial prey. Chinese perch were able to feed properly on live prey fish when either eyes or lateral lines were intact or functional, but could scarcely feed without these two senses. Chinese perch recognized its prey by vision through the perception of motion and shape, and showed a greater dependence on vision in predation when both visual and mechanical cues were available. Chemical stimulation by natural food could not elicit any feeding response in Chinese perch, and gustation was only important to the fish for the last stage of food discrimination in the oropharyngeal cavity. The sensory basis of Chinese perch in feeding is well adapted to its nocturnal stalking hunting strategy. and also explains its peculiar food habit of accepting live prey fish only and refusing dead prey fish or artificial diets. (C) 1998 The Fisheries Society of the British Isles.

基于手势的视觉用户界面研究

随着人机交互技术的发展，各种新的交互手段不断涌现，使人机交互朝着更加自然、高效和智能化的方向前进。基于手势的视觉用户界面是post-WIMP时代的一种重要的界面形式，与传统的WIMP交互方式相比视觉手势交互能够使用户摆脱鼠标键盘的束缚而采用一种更加自然、无约束的交互方式，从而提供给用户更大的交互空间、更多的交互自由度和更逼真的交互体验，具有较高的应用价值和良好的应用前景，因此被国内外越来越多的研究者所关注，迅速成为了人机交互领域一个热门的研究方向，并被广泛应用于虚拟/增强现实、普适计算、智能空间以及基于计算机的互动游戏等多个领域。 视觉交互是自然人机交互的核心和热点研究内容之一，视觉交互可以通过手势、目光、头部运动或者面部表情等多种方式进行。其中，手势是人类进行视觉交互的主要手段，手势所能表达的语义信息十分丰富。在人机交互中使用视觉手势完成交互任务不仅自然、直观和方便，而且从计算机的角度出发来看系统实现起来也较为容易。因此，基于手势的视觉用户界面得到了广泛的关注并取得了许多研究成果。从目前的研究现状来看，在基于手势的视觉用户界面研究中，仍然存在着以下的问题： （1）传统的WIMP界面模型已经不适合描述视觉手势界面这种post-WIMP界面的特点，需要对传统的用户界面模型加以扩展，构造适于描述视觉交互特征的界面模型帮助用户清晰、准确地分析和表达界面功能及其变化，描述出用户与系统的交互过程，指导软件系统的设计和实现； （2）具体的交互设计过程中还存在着许多关键技术没有解决好，例如手势的正确理解问题、算法的鲁棒性问题以及手势的可扩展性问题等； （3）传统视觉工具箱复杂难用的问题。需要降低开发难度，开发出便捷、易用、可扩展的软件界面开发工具及相应的开发方法，支持领域内非专家用户快速开发出视觉手势原型系统。 本文正是从以上问题出发，围绕着视觉手势界面交互技术，从理论、方法与应用等几个方面展开了深入的研究。首先论述了用户界面的发展历程，继而对视觉界面研究现状进行了综述。在充分研究和对比国内外相关研究成果的基础上提出了一种基于手势的视觉用户界面模型UIDT。接下来针对视觉手势交互中存在的难题，以认知心理学为理论依据提出了一种可扩展的手势交互状态转移模型，并在此基础上构建了视觉手势处理框架。随后对该框架中的关键技术进行了针对性的研究。针对非专家用户在构建具有个性化视觉手势界面过程中所遇到的问题，设计开发了一个支持视觉手势交互的开发工具IEToolkit并给出了一套基于该工具的通用的软件开发方法。最后，将上述成果应用于互动娱乐领域，取得了满意的效果。 本文的创新点主要表现在以下几个方面： 1. 提出了一种基于手势的视觉用户界面模型UIDT 在充分分析视觉交互特征的基础上，以传统的用户界面模型为基础提出了一种基于手势交互的视觉界面模型UIDT。该模型从用户模型、任务模型、设备模型和交互模型等几个方面对视觉手势交互进行了深入分析和描述，给出了各个模型的形式化定义，介绍了模型的各个组成模块以及它们之间的相互关系，讨论了视觉手势界面设计中应该遵循的设计规范。评估结果表明，该模型具有较强的通用性，有助于设计者对VBI的任务、用户、设备以及交互的不同层次进行抽象描述，使用户界面满足可用性要求，提高界面设计和原型开发的效率。 2. 提出了一个可扩展的视觉手势交互模型及一种新的视觉手势识别处理框架，围绕着该框架提出了一种新的视觉手势跟踪和识别方法 首先，提出了一个可扩展的视觉手势交互模型。根据认知心理学原理将视觉手势交互处理过程细分为选择性处理、分配性处理和集中处理三个不同的阶段，有效解决了Midas Touch问题；基于该模型提出了一个视觉手势识别框架，并结合认知心理学从手势检测、跟踪和识别三个方面对该框架的各个组成模块的关键技术进行了阐述。其中手势检测模块和识别管理模块能够辅助系统在复杂的背景中滤除掉不相关信息而选择性地搜索人手并根据上下文信息对手势识别任务重定向，从而避免了系统时刻都处于激活状态并对所有的手势动作都进行识别分析，有效解决了Midas Touch问题；为了提高系统的性能，提出了一种鲁棒的面向实时交互的变形手势跟踪方法和基于小样本学习的模板匹配方法用于动态手势识别。评估结果表明，上述技术有效地提高了系统的实时性、准确性和鲁棒性。 3. 设计开发了一个支持快速原型开发的视觉手势工具箱系统 针对目前大多数视觉手势工具箱复杂难用的问题，设计开发了一个简单、易用、可扩展的手势工具箱系统IEToolkit，实现了本文所述的各种关键交互技术。它包含了构造一个基于视觉手势的交互系统所需要的方方面面，从事件模型、交互模型、数据流模型等几个方面对工具箱的组成结构进行了描述，并给出了一个基于IEToolkit的通用的软件系统开发流程，开发人员可以将更多的精力集中在具体的高层逻辑语义处理上，而不需要过多考虑底层的技术细节与支撑结构。应用实例及评估结果表明IEToolkit能够降低开发门槛，较好地支持基于视觉手势交互的应用系统的快速构造，具有较高的应用价值。 4. 基于上述研究成果设计开发了一系列典型的视觉手势交互系统 将上述理论模型与具体的交互技术进行有效结合，在视觉手势开发工具平台基础上开发了一系列典型的互动娱乐系统，在实践中对本文的研究成果进行了有效性验证。