Barbara Villarini

Image quality assessment based on edge preservation

Objective image/video quality metrics which accurately represent the subjective quality of processed images are of paramount importance for the design and assessment of an image compression and transmission system. In some scenarios, it is also important to evaluate the quality of the received image with minimal reference to the transmitted one. For instance, for closed-loop optimization of a transmission system, the image quality measure can be evaluated at the receiver and provided as feedback information to the system controller. The original image - prior to compression and transmission - is not usually available at the receiver side, and it is important to rely at the receiver side on an objective quality metric that does not need reference or needs minimal reference to the original image. The observation that the human eye is very sensitive to edge and contour information of an image underpins the proposal of our reduced reference (RR) quality metric, which compares edge information between the distorted and the original image. Results highlight that the metric correlates well with subjective observations, also in comparison with commonly used full-reference metrics and with a state-of-the-art reduced reference metric.

A reduced-reference perceptual image and video quality metric based on edge preservation

In image and video compression and transmission, it is important to rely on an objective image/video quality metric which accurately represents the subjective quality of processed images and video sequences. In some scenarios, it is also important to evaluate the quality of the received video sequence with minimal reference to the transmitted one. For instance, for quality improvement of video transmission through closed-loop optimisation, the video quality measure can be evaluated at the receiver and provided as feedback information to the system controller. The original image/video sequence--prior to compression and transmission--is not usually available at the receiver side, and it is important to rely at the receiver side on an objective video quality metric that does not need reference or needs minimal reference to the original video sequence. The observation that the human eye is very sensitive to edge and contour information of an image underpins the proposal of our reduced reference (RR) quality metric, which compares edge information between the distorted and the original image. Results highlight that the metric correlates well with subjective observations, also in comparison with commonly used full-reference metrics and with a state-of-the-art RR metric.

A reprogrammable computing platform for JPEG 2000 and H.264 SHD video coding

In this paper, the architecture of a DSP/FPGA based hardware platform is presented, which is conceived to leverage programmable logic processing power for high definition video processing. The system is reconfigurable and scalable, since multiple boards may be parallelized to speed-up the most demanding tasks. JPEG 2000 and H.264, both at HD and Super HD (SHD) resolutions, have been simulated and their performance found on the embedded processing cores. The results show that real-time, or near real-time, encoding is viable, and the modularity of the architecture allows for parallelization and performance scalability.

MP33-20 THE SMARTTARGET BIOPSY TRIAL: A PROSPECTIVE PAIRED BLINDED TRIAL WITH RANDOMISATION TO COMPARE VISUAL-ESTIMATION AND IMAGE-FUSION TARGETED PROSTATE BIOPSIES

Multi-parametric MRI targeted prostate biopsies can improve detection of clinically significant prostate cancer and decrease the diagnosis of clinically insignificant cancers. There is debate whether visual estimated targeting is sufficient or whether image-fusion software is required. We conducted an ethics committee approved, prospective, blinded, paired validating clinical trial of visual estimated targeted biopsies compared to non-rigid MR/US image-fusion using an academically developed fusion system (SmartTarget®).

A Framework for Morphological Feature Extraction of Organs from MR Images for Detection and Classification of Abnormalities

In clinical practice, a misdiagnosis can lead to incorrect or delayed treatment, and in some cases, no treatment at all; consequently, the condition of a patient may worsen to varying degrees, in some cases proving fatal. The accurate 3D reconstruction of organs, which is a pioneering tool of medical image computing (MIC) technology, plays a key role in computer aided diagnosis (CADx), thereby enabling medical professionals to perform enhanced analysis on a region of interest. From here, the shape and structure of the organ coupled with measurements of its volume and curvature can provide significant guidance towards establishing the severity of a disorder or abnormality, consequently supporting improved diagnosis and treatment planning. Moreover, the classification and stratification of organ abnormalities is widely utilised within biomedical, forensic and MIC research for exploring and investigating organ deformations following injury, illness or trauma. This paper presents a tool that calculates, classifies and analyses pancreatic volume and curvature following their 3D reconstruction. Magnetic resonance imaging (MRI) volumes of 115 adult patients are evaluated in order to examine a correlation between these two variables. Such a tool can be utilised in the scope of much greater research and investigation. It can also be incorporated into the development of effective medical image analysis software application in the stratification of subjects and targeting of therapies.

An optimal method for searching UEP profiles in wireless JPEG 2000 video transmission

In this paper, we present a theoretical method to find an optimal source and channel code allocation, by considering an embedded image compression standard such as JPEG 2000 and a memoryless transmission channel. Lagrangian optimization is used to find an optimal error correction profile, from a PSNR perspective, starting from intermediate rate-distortion traces. The resulting mathematical relationship is simple and can be computed in real time. Simulations on a binary symmetric channel show that the achieved performance, in terms of PSNR, is comparable with that of similar methods reported in literature, while keeping a lower complexity.

Reduced-Reference Image Quality Assessment Based on Edge Preservation

Assessing the subjective quality of processed images through an objective quality metric is a key issue in multimedia processing and transmission. In some scenarios, it is also important to evaluate the quality of the received images with minimal reference to the transmitted ones. For instance, for closed-loop optimisation of image and video transmission, the quality measure can be evaluated at the receiver and provided as feedback information to the system controller. The original images - prior to compression and transmission - are not usually available at the receiver side, and it is important to rely at the receiver side on an objective quality metric that does not need reference or needs minimal reference to the original images. The observation that the human eye is very sensitive to edge and contour information of an image underpins the proposal of our reduced reference (RR) quality metric, which compares edge information between the distorted and the original image. Results highlight that the metric correlates well with subjective observations, also in comparison with commonly used full-reference metrics and with a state-of-the-art reduced reference metric.

Hierarchical Framework for Automatic Pancreas Segmentation in MRI Using Continuous Max-Flow and Min-Cuts Approach

Accurate, automatic and robust segmentation of the pancreas in medical image scans remains a challenging but important prerequisite for computer-aided diagnosis (CADx). This paper presents a tool for automatic pancreas segmentation in magnetic resonance imaging (MRI) scans. Proposed is a framework that employs a hierarchical pooling of information as follows: identify major pancreas region and apply contrast enhancement to differentiate between pancreatic and surrounding tissue; perform 3D segmentation by employing continuous max-flow and min-cuts approach, structured forest edge detection, and a training dataset of annotated pancreata; eliminate non-pancreatic contours from resultant segmentation via morphological operations on area, curvature and position between distinct contours. The proposed method is evaluated on a dataset of 20 MRI volumes, achieving a mean Dice Similarity coefficient of 75.5 ± 7.0% and a mean Jaccard Index coefficient of 61.2 ± 9.2%.

Cognitive behaviour analysis based on facial information using depth sensors

Cognitive behaviour analysis is considered of high importance with many innovative applications in a range of sectors including healthcare, education, robotics and entertainment. In healthcare, cognitive and emotional behaviour analysis helps to improve the quality of life of patients and their families. Amongst all the different approaches for cognitive behaviour analysis, significant work has been focused on emotion analysis through facial expressions using depth and EEG data. Our work introduces an emotion recognition approach using facial expressions based on depth data and landmarks. A novel dataset was created that triggers emotions from long or short term memories. This work uses novel features based on a non-linear dimensionality reduction, t-SNE, applied on facial landmarks and depth data. Its performance was evaluated in a comparative study, proving that our approach outperforms other state-of-the-art features.

Photometric Stereo for 3D Face Reconstruction Using Non Linear Illumination Models

Face recognition in presence of illumination changes, variant pose and different facial expressions is a challenging problem. In this paper, a method for 3D face reconstruction using photometric stereo and without knowing the illumination directions and facial expression is proposed in order to achieve improvement in face recognition. A dimensionality reduction method was introduced to represent the face deformations due to illumination variations and self shadows in a lower space. The obtained mapping function was used to determine the illumination direction of each input image and that direction was used to apply photometric stereo. Experiments with faces were performed in order to evaluate the performance of the proposed scheme. From the experiments it was shown that the proposed approach results very accurate 3D surfaces without knowing the light directions and with a very small differences compared to the case of known directions. As a result the proposed approach is more general and requires less restrictions enabling 3D face recognition methods to operate with less data.