Person tracking using motion detection and optical flow

Person tracking systems to date have either relied on motion detection or optical flow as a basis for person detection and tracking. As yet, systems have not been developed that utilise both these techniques. We propose a person tracking system that uses both, made possible by a novel hybrid optical flow-motion detection technique that we have developed. This provides the system with two methods of person detection, helping to avoid missed detections and the need to predict position, which can lead to errors in tracking and mistakes when handling occlusion situations. Our results show that our system is able to track people accurately, with an average error less than four pixels, and that our system outperforms the current CAVIAR benchmark system.

2D-3D Face Recognition Based on PCA and Feature Modelling

Hybrid face recognition, using image (2D) and structural (3D) information, has explored the fusion of Nearest Neighbour classifiers. This paper examines the effectiveness of feature modelling for each individual modality, 2D and 3D. Furthermore, it is demonstrated that the fusion of feature modelling techniques for the 2D and 3D modalities yields performance improvements over the individual classifiers. By fusing the feature modelling classifiers for each modality with equal weights the average Equal Error Rate improves from 12.60% for the 2D classifier and 12.10% for the 3D classifier to 7.38% for the Hybrid 2D+3D clasiffier.

Adaptive optical flow for person tracking

Person tracking systems are dependent on being able to locate a person accurately across a series of frames. Optical flow can be used to segment a moving object from a scene, provided the expected velocity of the moving object is known; but successful detection also relies on being able segment the background. A problem with existing optical flow techniques is that they don’t discriminate the foreground from the background, and so often detect motion (and thus the object) in the background. To overcome this problem, we propose a new optical flow technique, that is based upon an adaptive background segmentation technique, which only determines optical flow in regions of motion. This technique has been developed with a view to being used in surveillance systems, and our testing shows that for this application it is more effective than other standard optical flow techniques.

PhD forum : multiple camera management using wide base-line matching

Manual calibration of large and dynamic networks of cameras is labour intensive and time consuming. This is a strong motivator for the development of automatic calibration methods. Automatic calibration relies on the ability to find correspondences between multiple views of the same scene. If the cameras are sparsely placed, this can be a very difficult task. This PhD project focuses on the further development of uncalibrated wide baseline matching techniques.

Generation of a 3D proximal femur shape from a single projection 2D radiographic image.

Summary Generalized Procrustes analysis and thin plate splines were employed to create an average 3D shape template of the proximal femur that was warped to the size and shape of a single 2D radiographic image of a subject. Mean absolute depth errors are comparable with previous approaches utilising multiple 2D input projections. Introduction Several approaches have been adopted to derive volumetric density (g cm-3) from a conventional 2D representation of areal bone mineral density (BMD, g cm-2). Such approaches have generally aimed at deriving an average depth across the areal projection rather than creating a formal 3D shape of the bone. Methods Generalized Procrustes analysis and thin plate splines were employed to create an average 3D shape template of the proximal femur that was subsequently warped to suit the size and shape of a single 2D radiographic image of a subject. CT scans of excised human femora, 18 and 24 scanned at pixel resolutions of 1.08 mm and 0.674 mm, respectively, were equally split into training (created 3D shape template) and test cohorts. Results The mean absolute depth errors of 3.4 mm and 1.73 mm, respectively, for the two CT pixel sizes are comparable with previous approaches based upon multiple 2D input projections. Conclusions This technique has the potential to derive volumetric density from BMD and to facilitate 3D finite element analysis for prediction of the mechanical integrity of the proximal femur. It may further be applied to other anatomical bone sites such as the distal radius and lumbar spine.

Mid-level concept learning with visual contextual ontologies and probabilistic inference for image annotation

To date, automatic recognition of semantic information such as salient objects and mid-level concepts from images is a challenging task. Since real-world objects tend to exist in a context within their environment, the computer vision researchers have increasingly incorporated contextual information for improving object recognition. In this paper, we present a method to build a visual contextual ontology from salient objects descriptions for image annotation. The ontologies include not only partOf/kindOf relations, but also spatial and co-occurrence relations. A two-step image annotation algorithm is also proposed based on ontology relations and probabilistic inference. Different from most of the existing work, we specially exploit how to combine representation of ontology, contextual knowledge and probabilistic inference. The experiments show that image annotation results are improved in the LabelMe dataset.

Crowd Counting Using Multiple Local Features

In public venues, crowd size is a key indicator of crowd safety and stability. Crowding levels can be detected using holistic image features, however this requires a large amount of training data to capture the wide variations in crowd distribution. If a crowd counting algorithm is to be deployed across a large number of cameras, such a large and burdensome training requirement is far from ideal. In this paper we propose an approach that uses local features to count the number of people in each foreground blob segment, so that the total crowd estimate is the sum of the group sizes. This results in an approach that is scalable to crowd volumes not seen in the training data, and can be trained on a very small data set. As a local approach is used, the proposed algorithm can easily be used to estimate crowd density throughout different regions of the scene and be used in a multi-camera environment. A unique localised approach to ground truth annotation reduces the required training data is also presented, as a localised approach to crowd counting has different training requirements to a holistic one. Testing on a large pedestrian database compares the proposed technique to existing holistic techniques and demonstrates improved accuracy, and superior performance when test conditions are unseen in the training set, or a minimal training set is used.

Dense Correspondence Extraction in Difficult Uncalibrated Scenarios

The relationship between multiple cameras viewing the same scene may be discovered automatically by finding corresponding points in the two views and then solving for the camera geometry. In camera networks with sparsely placed cameras, low resolution cameras or in scenes with few distinguishable features it may be difficult to find a sufficient number of reliable correspondences from which to compute geometry. This paper presents a method for extracting a larger number of correspondences from an initial set of putative correspondences without any knowledge of the scene or camera geometry. The method may be used to increase the number of correspondences and make geometry computations possible in cases where existing methods have produced insufficient correspondences.

Semantic and context-based retrieval of digital cultural objects [Published in Chinese]

Cultural objects are increasingly generated and stored in digital form, yet effective methods for their indexing and retrieval still remain an important area of research. The main problem arises from the disconnection between the content-based indexing approach used by computer scientists and the description-based approach used by information scientists. There is also a lack of representational schemes that allow the alignment of the semantics and context with keywords and low-level features that can be automatically extracted from the content of these cultural objects. This paper presents an integrated approach to address these problems, taking advantage of both computer science and information science approaches. We firstly discuss the requirements from a number of perspectives: users, content providers, content managers and technical systems. We then present an overview of our system architecture and describe various techniques which underlie the major components of the system. These include: automatic object category detection; user-driven tagging; metadata transform and augmentation, and an expression language for digital cultural objects. In addition, we discuss our experience on testing and evaluating some existing collections, analyse the difficulties encountered and propose ways to address these problems.

Assessment of tear film surface quality using dynamic-area high-speed videokeratoscopy

A new method for noninvasive assessment of tear film surface quality (TFSQ) is proposed. The method is based on high-speed videokeratoscopy in which the corneal area for the analysis is dynamically estimated in a manner that removes videokeratoscopy interference from the shadows of eyelashes but not that related to the poor quality of the precorneal tear film that is of interest. The separation between the two types of seemingly similar videokeratoscopy interference is achieved by region-based classification in which the overall noise is first separated from the useful signal (unaltered videokeratoscopy pattern), followed by a dedicated interference classification algorithm that distinguishes between the two considered interferences. The proposed technique provides a much wider corneal area for the analysis of TFSQ than the previously reported techniques. A preliminary study with the proposed technique, carried out for a range of anterior eye conditions, showed an effective behavior in terms of noise to signal separation, interference classification, as well as consistent TFSQ results. Subsequently, the method proved to be able to not only discriminate between the bare eye and the lens on eye conditions but also to have the potential to discriminate between the two types of contact lenses.

Design and evaluation of an image analysis platform for low-power, low-bandwidth camera networks

We describe the design and evaluation of a platform for networks of cameras in low-bandwidth, low-power sensor networks. In our work to date we have investigated two different DSP hardware/software platforms for undertaking the tasks of compression and object detection and tracking. We compare the relative merits of each of the hardware and software platforms in terms of both performance and energy consumption. Finally we discuss what we believe are the ongoing research questions for image processing in WSNs.