RESUMEN
Camera-based object detection is integral to advanced driver assistance systems (ADAS) and autonomous vehicle research, and RGB cameras remain indispensable for their spatial resolution and color information. This study investigates exposure time optimization for such cameras, considering image quality in dynamic ADAS scenarios. Exposure time, the period during which the camera sensor is exposed to light, directly influences the amount of information captured. In dynamic scenarios, such as those encountered in typical driving scenarios, optimizing exposure time becomes challenging due to the inherent trade-off between Signal-to-Noise Ratio (SNR) and motion blur, i.e., extending exposure time to maximize information capture increases SNR, but also increases the risk of motion blur and overexposure, particularly in low-light conditions where objects may not be fully illuminated. The study introduces a comprehensive methodology for exposure time optimization under various lighting conditions, examining its impact on image quality and computer vision performance. Traditional image quality metrics show a poor correlation with computer vision performance, highlighting the need for newer metrics that demonstrate improved correlation. The research presented in this paper offers guidance into the enhancement of single-exposure camera-based systems for automotive applications. By addressing the balance between exposure time, image quality, and computer vision performance, the findings provide a road map for optimizing camera settings for ADAS and autonomous driving technologies, contributing to safety and performance advancements in the automotive landscape.
RESUMEN
In recent years, significant advances have been made in the development of Advanced Driver Assistance Systems (ADAS) and other technology for autonomous vehicles. Automated object detection is a crucial component of autonomous driving; however, there are still known issues that affect its performance. For automotive applications, object detection algorithms are required to perform at a high standard in all lighting conditions; however, a major problem for object detection is poor performance in low-light conditions due to objects being less visible. This study considers the impact of training data composition on object detection performance in low-light conditions. In particular, this study evaluates the effect of different combinations of images of outdoor scenes, from different times of day, on the performance of deep neural networks, and considers the different challenges encountered during the training of a neural network. Through experiments with a widely used public database, as well as a number of commonly used object detection architectures, we show that more robust performance can be obtained with an appropriate balance of classes and illumination levels in the training data. The results also highlight the potential of adding images obtained in dusk and dawn conditions for improving object detection performance in day and night.
RESUMEN
Image decolorization is an image pre-processing step which is widely used in image analysis, computer vision, and printing applications. The most commonly used methods give each color channel (e.g., the R component in RGB format, or the Y component of an image in CIE-XYZ format) a constant weight without considering image content. This approach is simple and fast, but it may cause significant information loss when images contain too many isoluminant colors. In this paper, we propose a new method which is not only efficient, but also can preserve a higher level of image contrast and detail than the traditional methods. It uses the information from the cumulative distribution function (CDF) of the information in each color channel to compute a weight for each pixel in each color channel. Then, these weights are used to combine the three color channels (red, green, and blue) to obtain the final grayscale value. The algorithm works in RGB color space directly without any color conversion. In order to evaluate the proposed algorithm objectively, two new metrics are also developed. Experimental results show that the proposed algorithm can run as efficiently as the traditional methods and obtain the best overall performance across four different metrics.
RESUMEN
In advanced driver assistance systems (ADAS) or autonomous vehicle research, acquiring semantic information about the surrounding environment generally relies heavily on camera-based object detection. Image signal processors (ISPs) in cameras are generally tuned for human perception. In most cases, ISP parameters are selected subjectively and the resulting image differs depending on the individual who tuned it. While the installation of cameras on cars started as a means of providing a view of the vehicle's environment to the driver, cameras are increasingly becoming part of safety-critical object detection systems for ADAS. Deep learning-based object detection has become prominent, but the effect of varying the ISP parameters has an unknown performance impact. In this study, we analyze the performance of 14 popular object detection models in the context of changes in the ISP parameters. We consider eight ISP blocks: demosaicing, gamma, denoising, edge enhancement, local tone mapping, saturation, contrast, and hue angle. We investigate two raw datasets, PASCALRAW and a custom raw dataset collected from an advanced driver assistance system (ADAS) perspective. We found that varying from a default ISP degrades the object detection performance and that the models differ in sensitivity to varying ISP parameters. Finally, we propose a novel methodology that increases object detection model robustness via ISP variation data augmentation.
RESUMEN
Although cochlear implants work well for people with hearing impairment in quiet conditions, it is well-known that they are not as effective in noisy environments. Noise reduction algorithms based on machine learning allied with appropriate speech features can be used to address this problem. The purpose of this study is to investigate the importance of acoustic features in such algorithms. Acoustic features are extracted from speech and noise mixtures and used in conjunction with the ideal binary mask to train a deep neural network to estimate masks for speech synthesis to produce enhanced speech. The intelligibility of this speech is objectively measured using metrics such as Short-time Objective Intelligibility (STOI), Hit Rate minus False Alarm Rate (HIT-FA) and Normalized Covariance Measure (NCM) for both simulated normal-hearing and hearing-impaired scenarios. A wide range of existing features is experimentally evaluated, including features that have not been traditionally applied in this application. The results demonstrate that frequency domain features perform best. In particular, Gammatone features performed best for normal hearing over a range of signal-to-noise ratios and noise types (STOI = 0.7826). Mel spectrogram features exhibited the best overall performance for hearing impairment (NCM = 0.7314). There is a stronger correlation between STOI and NCM than HIT-FA and NCM, suggesting that the former is a better predictor of intelligibility for hearing-impaired listeners. The results of this study may be useful in the design of adaptive intelligibility enhancement systems for cochlear implants based on both the noise level and the nature of the noise (stationary or non-stationary).
Asunto(s)
Implantación Coclear , Implantes Cocleares , Humanos , Acústica , Algoritmos , BenchmarkingRESUMEN
Interacting with other roads users is a challenge for an autonomous vehicle, particularly in urban areas. Existing vehicle systems behave in a reactive manner, warning the driver or applying the brakes when the pedestrian is already in front of the vehicle. The ability to anticipate a pedestrian's crossing intention ahead of time will result in safer roads and smoother vehicle maneuvers. The problem of crossing intent forecasting at intersections is formulated in this paper as a classification task. A model that predicts pedestrian crossing behaviour at different locations around an urban intersection is proposed. The model not only provides a classification label (e.g., crossing, not-crossing), but a quantitative confidence level (i.e., probability). The training and evaluation are carried out using naturalistic trajectories provided by a publicly available dataset recorded from a drone. Results show that the model is able to predict crossing intention within a 3-s time window.
RESUMEN
Cochlear implant technology successfully restores hearing function to patients with sensory impairment. Although cochlear implant users generally hear well in quiet, they still find noisy conditions very challenging, hence the need to employ noise reduction algorithms in these systems to enhance the user experience. This paper reviews noise reduction algorithms in cochlear implants. Traditionally, such algorithms have been classified as either single- or multiple-channel, depending on the number of microphones they use. This review retains this general classification in looking at recent papers and extends it to reflect recent interest in machine learning techniques. The review concludes with consideration of promising future areas of research.
Asunto(s)
Implantación Coclear , Implantes Cocleares , Percepción del Habla , Humanos , Implantación Coclear/métodos , Ruido , Procesamiento de Señales Asistido por ComputadorRESUMEN
Paroxysmal atrial fibrillation (PAF) is a cardiac arrhythmia that can eventually lead to heart failure or stroke if left untreated. Early detection of PAF is therefore crucial to prevent any further complications and avoid fatalities. An implantable defibrillator device could be used to both detect and treat the condition though such devices have limited computational capability. With this constraint in mind, this paper presents a novel set of features to accurately predict the presence of PAF. The method is evaluated using ECG signals from the widely used atrial fibrillation prediction database (AFPDB) from PhysioNet. We analysed 106 signals from 53 pairs of ECG recordings. Each pair of signals contains one 5-min ECG segment that ends just before the onset of a PAF event and another 5-min ECG segment at least 45 min distant from the PAF event, to represent a non-PAF event. Seven novel features are extracted through the Poincaré representation of R-R interval signals, and are prioritised through feature ranking schemes. The features are used with four standard classification techniques for PAF prediction and compared to the existing state of the art from the literature. Using only the seven proposed features, classification performance outperforms those of the classical state-of-the-art feature set, registering sensitivity and specificity measurements of over 96%. The results further improve when the features are combined with several of the classical features, with an accuracy increasing to 98% using a linear kernel SVM. The results show that the proposed features provide a useful representation of the PAF condition and achieve good prediction with off-the-shelf classification techniques that would be suitable for ICU deployment.
Asunto(s)
Fibrilación Atrial , Algoritmos , Fibrilación Atrial/diagnóstico , Bases de Datos Factuales , Electrocardiografía , Frecuencia Cardíaca , HumanosRESUMEN
Over the last four decades, implantable cardioverter defibrillators (ICDs) have been widely deployed to reduce sudden cardiac death (SCD) risk in patients with a history of life-threatening arrhythmia. By continuous monitoring of the heart rate, ICDs can use decision algorithms to distinguish normal cardiac sinus rhythm or supra-ventricular tachycardia from abnormal cardiac rhythms like ventricular tachycardia and ventricular fibrillation and deliver appropriate therapy such as an electrical stimulus. Despite the success of ICDs, more research is still needed, particularly in decision-making algorithms. Because of low specificity in practical devices, patients with ICDs still receive inappropriate shocks, which may lead to inadvertent mortality and reduction of quality of life. At the same time, higher sensitivity can lead to the use of newer tiered therapies. The purpose of this study is to review the literature on common signal features used in detection algorithms for abnormal cardiac sinus rhythm, as well as reviewing datasets used for algorithm development in previous studies. More than 50 different features to address heart rate changes before SCD have been reviewed and general methodology on this area proposed based on variety of studies on ICDs functionality. A comparative study on the prediction performance of these features, using a common database, is also presented. By combining these features with a support vector machine classifier, achieved results have compared well with other studies.
Asunto(s)
Arritmias Cardíacas , Muerte Súbita Cardíaca , Desfibriladores Implantables , Procesamiento de Señales Asistido por Computador , Adulto , Anciano , Algoritmos , Arritmias Cardíacas/diagnóstico , Arritmias Cardíacas/fisiopatología , Muerte Súbita Cardíaca/epidemiología , Muerte Súbita Cardíaca/prevención & control , Femenino , Frecuencia Cardíaca/fisiología , Humanos , Masculino , Persona de Mediana Edad , Adulto JovenRESUMEN
Radar-based breast imaging has shown promise as an imaging modality for early-stage cancer detection, and clinical investigations of two commercial imaging systems are ongoing. Many imaging algorithms have been proposed, which seek to improve the quality of the reconstructed microwave image to enhance the potential clinical decision. However, in many cases, the radar-based imaging algorithms have only been tested in limited numerical or experimental test cases or with simplifying assumptions such as using one estimate of permittivity for all patient test cases. In this work, the potential impact of patient-specific permittivity estimation on algorithm comparison is highlighted using representative experimental breast phantoms. In particular, the case studies presented help show that the permittivity estimate can impact the conclusions of the algorithm comparison. Overall, this work suggests that it is important that imaging algorithm comparisons use realistic test cases with and without breast abnormalities and with reconstruction permittivity estimation.
RESUMEN
Many new clinical investigations of microwave breast imaging have been published in recent years. Trials with over one hundred participants have indicated the potential of microwave imaging to detect breast cancer, with particularly encouraging sensitivity results reported from women with dense breasts. The next phase of clinical trials will involve larger and more diverse populations, including women with no breast abnormalities or benign breast diseases. These trials will need to address clinical efficacy in terms of sensitivity and specificity. A number of challenges exist when using microwave imaging with broad populations: 1) addressing the substantial variance in breast composition observed in the population and 2) achieving high specificity given differences between individuals. This paper analyses these challenges using a diverse phantom set which models the variance in breast composition and tumor shape and size seen in the population. The data show that the sensitivity of microwave breast imaging in breasts of differing density can suffer if patient-specific beamforming is not used. Moreover, the results suggest that achieving high specificity in dense breasts may be difficult, but that patient-specific beamforming does not adversely affect the expected specificity. In summary, this paper finds that patient-specific beamforming has a tangible impact on expected sensitivity in experimental cases and that achieving high specificity in dense breasts may be challenging.
Asunto(s)
Mama/diagnóstico por imagen , Procesamiento de Imagen Asistido por Computador/métodos , Imágenes de Microonda , Algoritmos , Neoplasias de la Mama/diagnóstico por imagen , Femenino , Humanos , Fantasmas de ImagenRESUMEN
Implantable cardioverter defibrillators (ICDs) are commonly used in patients at high risk of sudden cardiac death (SCD) to help prevent and treat life-threatening arrhythmia. Up to 80% of cases of sudden cardiac death are caused by ventricular tachyarrhythmias (VTA) and the accurate prediction of VTA in patients with ICDs can help prevent SCD. Early prediction allows tiered and less invasive therapies to be used to help prevent VTA which are more easily tolerated by the patient and are less battery intensive. In this work, a comparative study of three types of frequency domain features (spectral, bispectrum, and Fourier-Bessel) for VTA prediction is presented based on heart rate variability (HRV) signals between one and five minutes prior to known SCD. Using Fourier-Bessel features and a standard classification approach resulted in the best performance of 87.5% accuracy, 89.3% sensitivity and 85.7% specificity. These results suggest that Fourier-Bessel features are a promising approach for SCD prediction, and that new feature development can help improve both the sensitivity and specificity of SCD prediction in ICDs.
Asunto(s)
Desfibriladores Implantables , Taquicardia Ventricular , Arritmias Cardíacas , Muerte Súbita Cardíaca , Frecuencia Cardíaca , Humanos , TaquicardiaRESUMEN
Microwave breast imaging has seen significant developments in recent years, including new clinical trials and formation of a number of spin-out companies. Although many algorithms for microwave breast imaging have been developed, there are significant challenges in translating these algorithms to the clinic. For example, movement due to patient breathing can affect the scan, and both the breast and breast abnormalities vary significantly from patient to patient. As breast density is a known independent risk factor for cancer and cancerous tumours have different shapes and margins to benign tumours, the effect of interpatient variance on the microwave image is important. This work analyses the effect on image quality of tumour shape, size and breast density. Using the diverse and representative BRIGID experimental dataset, images of a variety of tumours are compared to images without tumours present. This work suggests that it is difficult to distinguish images with and without tumours present using existing metrics.
Asunto(s)
Microondas , Algoritmos , Mama , Densidad de la Mama , Neoplasias de la Mama , HumanosRESUMEN
Determination of the sludge volume index is key to describing the settling characteristics of sludge in the aeration process of wastewater treatment plants (WWTPs). The two core components of this calculation are the settled sludge volume (SSV) and suspended solids. While the measurement procedure for SSV is generally defined by national or international standards, in practice a wide variety of vessel sizes and shapes are used by operators to monitor WWTP performance. Furthermore, differences in how these tests are carried out can lead to poor data, inefficient WWTP operation and a lack of comparable metrics for WWTP operational monitoring. Thus, there is a requirement to improve operational performance of WWTPs to meet the increasingly stringent legislation regarding discharge limits. The aim of this study was to utilise a novel image-processing system (AutoSSV) to (i) determine its efficacy in describing SSV and (ii) measure and compare different methodologies for measurement of SSV. The AutoSSV system was tested using samples from various WWTPs and the results compared to those determined by standard manual measurement. Both standard and modified settlement tests were conducted on 30 mixed liquor samples, with modified settlement tests consistently resulting in lower SSV measurements. Results from the study showed a strong correlation between the SSV measurements provided by the AutoSSV system and results obtained from current manual measurement methods. The proposed technique would help to standardise the measurement in practice and increase the frequency of monitoring, particularly in small-scale rural WWTPs where there may not be permanent operators on site, and thus provide sufficient performance monitoring for efficient and effective operation.
Asunto(s)
Procesamiento de Imagen Asistido por Computador/métodos , Aguas del Alcantarillado/análisis , Eliminación de Residuos Líquidos/métodosRESUMEN
OBJECTIVE: Microwave breast imaging has seen significant academic and commercial development in recent years, with four new operational microwave imaging systems used with patients since 2016. In this paper, a comprehensive review of these recent clinical advances is presented, comparing patient populations and study outcomes. For the first time, the designs of operational microwave imaging systems are compared in detail. METHODS: First, the current understanding of dielectric properties of human breast tissues is reviewed, considering evidence from operational microwave imaging systems and from dielectric properties measurement studies. Second, design features of operational microwave imaging systems are discussed in terms of advantages and disadvantages during clinical operation. RESULTS: Reported results from patient imaging trials are compared, contrasting the principal results from each trial. Additionally, clinical experience from each trial is highlighted, identifying desirable system design features for clinical use. CONCLUSIONS: Increasingly, evidence from patient imaging studies indicate that a contrast in dielectric properties between healthy and cancerous breast tissues exists. However, despite the significant and encouraging results from patient trials, variation still exists in the microwave imaging system design. SIGNIFICANCE: This study seeks to define the current state of the art in microwave breast imaging, and identify suitable design characteristics for ease of clinical use.
Asunto(s)
Mama/diagnóstico por imagen , Interpretación de Imagen Asistida por Computador/métodos , Microondas/uso terapéutico , Adulto , Anciano , Anciano de 80 o más Años , Algoritmos , Neoplasias de la Mama/diagnóstico por imagen , Diagnóstico por Imagen/métodos , Femenino , Humanos , Persona de Mediana Edad , Fantasmas de ImagenRESUMEN
Confocal Microwave Imaging (CMI) for the early detection of breast cancer has been under development for over two decades and is currently going through early-phase clinical evaluation. The image reconstruction algorithm is a key signal processing component of any CMI-based breast imaging system and impacts the efficacy of CMI in detecting breast cancer. Several image reconstruction algorithms for CMI have been developed since its inception. These image reconstruction algorithms have been previously evaluated and compared, using both numerical and physical breast models, and healthy volunteer data. However, no study has been performed to evaluate the performance of image reconstruction algorithms using clinical patient data. In this study, a variety of imaging algorithms, including both data-independent and data-adaptive algorithms, were evaluated using data obtained from a small-scale patient study conducted at the University of Calgary. Six imaging algorithms were applied to reconstruct 3D images of five clinical patients. Reconstructed images for each algorithm and each patient were compared to the available clinical reports, in terms of abnormality detection and localisation. The imaging quality of each algorithm was evaluated using appropriate quality metrics. The results of the conventional Delay-and-Sum algorithm and the Delay-Multiply-and-Sum (DMAS) algorithm were found to be consistent with the clinical information, with DMAS producing better quality images compared to all other algorithms.
Asunto(s)
Algoritmos , Procesamiento de Imagen Asistido por Computador , Imagenología Tridimensional , Microondas , Pacientes , Procesamiento de Señales Asistido por Computador , HumanosRESUMEN
Currently, breast cancer often requires invasive biopsies for diagnosis, motivating researchers to design and develop non-invasive and automated diagnosis systems. Recent microwave breast imaging studies have shown how backscattered signals carry relevant information about the shape of a tumour, and tumour shape is often used with current imaging modalities to assess malignancy. This paper presents a comprehensive analysis of microwave breast diagnosis systems which use machine learning to learn characteristics of benign and malignant tumours. The state-of-the-art, the main challenges still to overcome and potential solutions are outlined. Specifically, this work investigates the benefit of signal pre-processing on diagnostic performance, and proposes a new set of extracted features that capture the tumour shape information embedded in a signal. This work also investigates if a relationship exists between the antenna topology in a microwave system and diagnostic performance. Finally, a careful machine learning validation methodology is implemented to guarantee the robustness of the results and the accuracy of performance evaluation.
RESUMEN
Wastewater treatment facilities are continually challenged to meet both environmental regulations and reduce running costs (particularly energy and staffing costs). Improving the efficiency of operational monitoring at wastewater treatment plants (WWTPs) requires the development and implementation of appropriate performance metrics; particularly those that are easily measured, strongly correlate to WWTP performance, and can be easily automated, with a minimal amount of maintenance or intervention by human operators. Turbidity is the measure of the relative clarity of a fluid. It is an expression of the optical property that causes light to be scattered and absorbed by fine particles in suspension (rather than transmitted with no change in direction or flux level through a fluid sample). In wastewater treatment, turbidity is often used as an indicator of effluent quality, rather than an absolute performance metric, although correlations have been found between turbidity and suspended solids. Existing laboratory-based methods to measure turbidity for WWTPs, while relatively simple, require human intervention and are labour intensive. Automated systems for on-site measuring of wastewater effluent turbidity are not commonly used, while those present are largely based on submerged sensors that require regular cleaning and calibration due to fouling from particulate matter in fluids. This paper presents a novel, automated system for estimating fluid turbidity. Effluent samples are imaged such that the light absorption characteristic is highlighted as a function of fluid depth, and computer vision processing techniques are used to quantify this characteristic. Results from the proposed system were compared with results from established laboratory-based methods and were found to be comparable. Tests were conducted using both synthetic dairy wastewater and effluent from multiple WWTPs, both municipal and industrial. This system has an advantage over current methods as it provides a multipoint analysis that can be easily repeated for large volumes of wastewater effluent. Although the system was specifically designed and tested for wastewater treatment applications, it could have applications such as in drinking water treatment, and in other areas where fluid turbidity is an important measurement.
Asunto(s)
Procesamiento de Imagen Asistido por Computador/métodos , Eliminación de Residuos Líquidos/métodos , Aguas Residuales/química , Composición Familiar , Procesamiento de Imagen Asistido por Computador/instrumentación , Residuos Industriales , Purificación del Agua/métodosRESUMEN
Inaccurate estimation of average dielectric properties can have a tangible impact on microwave radar-based breast images. Despite this, recent patient imaging studies have used a fixed estimate although this is known to vary from patient to patient. Parameter search algorithms are a promising technique for estimating the average dielectric properties from the reconstructed microwave images themselves without additional hardware. In this work, qualities of accurately reconstructed images are identified from point spread functions. As the qualities of accurately reconstructed microwave images are similar to the qualities of focused microscopic and photographic images, this work proposes the use of focal quality metrics for average dielectric property estimation. The robustness of the parameter search is evaluated using experimental dielectrically heterogeneous phantoms on the three-dimensional volumetric image. Based on a very broad initial estimate of the average dielectric properties, this paper shows how these metrics can be used as suitable fitness functions in parameter search algorithms to reconstruct clear and focused microwave radar images.
Asunto(s)
Microondas , Algoritmos , Mama , Neoplasias de la Mama , Humanos , Fantasmas de Imagen , RadarRESUMEN
This paper proposes a novel adaptive dictionary (AD) reconstruction scheme to improve the performance of compressed sensing (CS) with electrocardiogram signals (ECG). The method is based on the use of multiple dictionaries, created using dictionary learning (DL) techniques for CS signal reconstruction. The modified reconstruction framework is a two-stage process that leverages information about the signal from an initial signal reconstruction stage. By identifying whether a QRS complex is present and if so, determining a location estimate of the QRS, the most appropriate dictionary is selected and a second stage more refined signal reconstruction can be obtained. The performance of the proposed algorithm is compared with state-of-the-art CS implementations in the literature, as well as the set partitioning in hierarchical trees (SPIHT) wavelet-based lossy compression algorithm. The results indicate that the proposed reconstruction scheme outperforms all existing CS implementations in terms of signal fidelity at each compression ratio tested. The performance of the proposed approach also compares favorably with SPIHT in terms of signal reconstruction quality. Furthermore, an analysis of the overall power consumption of the proposed ECG compression framework as would be used in a body area network (BAN) demonstrates positive results for the proposed CS approach when compared with existing CS techniques and SPIHT.