In traditional industry, the detection of the automobile components quality is mainly completed by human eyes. Low detection accuracy, high labor consumption, and slow detection speed are important reasons for the slo...
详细信息
Heart rate is a crucial metric in health monitoring. Traditional computer vision solutions estimate cardiac signals by detecting physical manifestations of heartbeats, such as facial discoloration caused by blood oxyg...
详细信息
Heart rate is a crucial metric in health monitoring. Traditional computer vision solutions estimate cardiac signals by detecting physical manifestations of heartbeats, such as facial discoloration caused by blood oxygenation changes, from subject videos using regression methods. As continuous signals are more complex and expensive to de-noise, this study introduces an alternative approach, employing end-to-end classification models to remotely derive a discrete representation of cardiac signals from face videos. These visual cardiac signal classifiers are trained on discretized cardiac signals, a novel pre-processing method with limited precedent in health monitoring literature. Consequently, various methods to convert continuous cardiac signals into binary form are presented, and their impact on training is evaluated. An implementation of this approach, the temporal shift convolutional attention binary classifier, is presented using the regression-based convolutional attention network architecture. The classifier and a baseline regression model are trained and tested using publicly available and locally collected datasets designed for heart signal detection from face video. The model performance is then assessed based on the heart rate error from the extracted cardiac signals. Results show the proposed method outperforms the baseline on the UBFC-rPPG dataset, reducing cross-dataset root mean square error from 2.33 to 1.63 beats per minute. However, both models struggled to generalize to the PURE dataset, with root mean square errors of 12.40 and 16.29 beats per minute, respectively. Additionally, the proposed approach reduces the computational complexity of model output post-processing, enhancing its suitability for real-time applications and deployment on systems with restricted resources.
Leaf wetness duration is a crucial factor in plant disease management. Current optical methods use standard RGB images to classify leaf wetness as a binary problem, i.e., wet or dry. Green leaves absorb red light, whe...
详细信息
Leaf wetness duration is a crucial factor in plant disease management. Current optical methods use standard RGB images to classify leaf wetness as a binary problem, i.e., wet or dry. Green leaves absorb red light, whereas water reflects it. Based on this difference, an experimental platform was built to semi-automatically measure droplet deposition on grape leaves while capturing red laser images using an RGB camera. The setup measured changes in leaf mass and area of scanned leaves to determine the water mass per leaf area as a measure of leaf wetness. A sprayer was used to apply water droplets to the leaves. As the amount of deposited water increased, the mean red channel intensity decreased, with more bright spots in the images. These bright spots were more distinguishable as droplets in the green channel. Segmented leaf area, mean red channel intensity, and the number of identified droplets were used as image features. A generalised additive model was employed to predict the leaf wetness value with extracted features. The R-squared value for the prediction of the validation dataset was 0.71. image resolution and leaf orientation were identified as factors that influenced the model accuracy. The measurement method introduced in this study shows potential for accurately quantifying leaf wetness, and implies that in practice detecting leaf wetness can be integrated into a multi-classification problem, thereby broadening the potential applications of optical methods.
Depth image spatial clustering is an important task in the fields of computer vision and machine learning, aiming to group pixels or point cloud data of depth images into clusters with similar features. This is crucia...
详细信息
Diabetes-oriented diabetic retinopathy impacts the blood vessels in the region of the retina to enlarge and leak blood and other fluids. In most cases, diabetic retinopathy affects both eyes. If left untreated, it wou...
详细信息
machinevision technology has the characteristics of high precision, fast speed, non-contact and high degree of automation in displacement detection. It is feasible to apply machinevision technology to the detection ...
详细信息
In this paper, a computer vision-based approach for optimizing component test benches in endurance testing of automotive components. As a use case, the paper explores testing of automotive throttle position sensor usi...
详细信息
One of the most widely researched topics in the food industry is bread quality analysis. Different techniques have been developed to assess the quality characteristics of bakery products. However, in the last few deca...
详细信息
One of the most widely researched topics in the food industry is bread quality analysis. Different techniques have been developed to assess the quality characteristics of bakery products. However, in the last few decades, the advancement in sensor and computational technologies has increased the use of computer vision to analyze food quality (e.g., bakery products). Despite a large number of publications on the application of imaging methods in the bakery industry, comprehensive reviews detailing the use of conventional analytical techniques and imaging methods for the quality analysis of baked goods are limited. Therefore, this review aims to critically analyze the conventional methods and explore the potential of imaging techniques for the quality assessment of baked products. This review provides an in-depth assessment of the different conventional techniques used for the quality analysis of baked goods which include methods to record the physical characteristics of bread and analyze its quality, sensory-based methods, nutritional-based methods, and the use of dough rheological data for end-product quality prediction. Furthermore, an overview of the imageprocessing stages is presented herein. We also discuss, comprehensively, the applications of imaging techniques for assessing the quality of bread and other baked goods. These applications include studying and predicting baked goods' quality characteristics (color, texture, size, and shape) and classifying them based on these features. The limitations of both conventional techniques (e.g., destructive, laborious, error-prone, and expensive) and imaging methods (e.g., illumination, humidity, and noise) and the future direction of the use of imaging methods for quality analysis of bakery products are discussed.
The brain tumor is formed by abnormal cells that develop and reproduce unpredictably. A timely diagnosis of a brain tumor amplifies the likelihood of living for the patient. Specialists generally deploy a manual metho...
详细信息
The brain tumor is formed by abnormal cells that develop and reproduce unpredictably. A timely diagnosis of a brain tumor amplifies the likelihood of living for the patient. Specialists generally deploy a manual methodology of segmentation when diagnosing brain tumours. In medical imageprocessing, brain tumour fragmentation is significant. The Physicians typically employed a manual process of fragmentation when identifying brain tumours. It is not exact, is subject to inter-and intra-observer variability, and may include non-enhancing tissue. It is also time demanding. A new and Improved Deep Learning Model formulated on the Cascade Regression method (DLCR) is proposed for image segmentation to resolve these issues. The proposed method uses the normalization procedure for Pre-processing of Magnetic resonance imaging (MRI) images using Fully Convolutional Neural Network (FCNN) method. Then the feature extraction using the Gaussian Mixture Model (GMM) is utilized to to decrease the data and obtain the relevant characteristic from every feature vector. Then the Current methodologies, namely machine Learning Predictive Model (MLPM), Deep Learning Framework (DLF) and Extreme Learning machine Local Receptive Fields (ELM-LRF) were compared to our suggested method. The results show the proposed DLCR method has achieved a better sensitivity, specificity, recall ratio, precision ratio, peak signal-to-noise ratio (PSNR), and low Root Mean Square Error (RMSE) than the existing methods.
Imaging systems work diversely in the imageprocessing domain, and each system contains specific characteristics. We are developing models to fuse images from different sensors and environments to get promising outcom...
详细信息
ISBN:
(纸本)9798400709234
Imaging systems work diversely in the imageprocessing domain, and each system contains specific characteristics. We are developing models to fuse images from different sensors and environments to get promising outcomes for different computer visionapplications. The multiple unified models have been developed for multiple tasks such as multi-focus (MF), multi-exposure (ME), and multimodal (MM) image fusion. The careful tuning of such models is required to get optimal results, which are still not applicable to diverse applications. We propose an automatic machine learning (AML) based multi-tasking image fusion approach to overcome this problem. Initially, we evaluate source images with AML and feed them to the task-based models. Then, the source images are fused with the pre-trained and fine-tuned models. The experimental results authenticate the consequences of our proposed approach compared to generic approaches.
暂无评论