Explore the vast range of titles available.

With the continuous development of remote sensing technology, more and more image data have been acquired. How to store these data safely and effectively has become an urgent problem that needs to be solved. In this paper, after analysing the types and development methods of image data, we decided to use VC to develop based on Oracle's image database; an image database system is designed and implemented. Specific research work is mainly reflected in many aspects:

Diabetic retinopathy (DR), a prevalent microvascular complication of diabetes, is the fifth leading cause of blindness worldwide. Given the critical nature of the disease, it is paramount that individuals with diabetes undergo annual screening for early and timely detection of DR, facilitating prompt ophthalmic assessment and intervention. However, screening for DR, which involves assessing visual acuity and retinal examination through ophthalmoscopy or retinal photography, presents a significant global challenge due to the massive volume of individuals requiring annual reviews. To counter this challenge, there has been an increasing interest in the potential of artificial intelligence (AI) tools for automated diagnosis of DR. The AI tools primarily utilize deep learning (DL) techniques and are tailored to analyse extensive medical image data and provide diagnostic outputs, essentially streamline the DR screening process. However, the development of such AI tools requires access to a comprehensive retinal image database with a plethora of high-resolution fundus images from various cameras, covering all DR lesions. Additionally, the accurate training of these AI algorithms necessitates skilled professionals, such as optometrists or ophthalmologists, to provide reliable ground truths that ensure the precision of the diagnostic outputs. To address these prerequisites, we have initiated a study involving multiple institutions to establish a large-scale online 'Retinal Image Database’ in India, aiming to contribute significantly to DR research. This paper delineates the methodology employed for this significant undertaking, detailing the steps taken to create the large retinal image database, as well as the framework for developing a cost-effective, robust AI-based DR diagnostic tool. Our work is expected to mark a significant stride in DR detection and management, promising a more efficient and scalable solution for tackling this global health challenge.

Subjective image quality databases are a major source of raw data on how the visual system works in . These databases describe the sensitivity of many observers to a wide range of distortions of different nature and intensity seen on top of a variety of natural images. Data of this kind seems to open a number of possibilities for the vision scientist to check the models in realistic scenarios. However, while these natural databases are great benchmarks for models developed in some other way (e.g., by using the well-controlled of traditional psychophysics), they should be carefully used when trying to fit vision models. Given the high dimensionality of the image space, it is very likely that some basic phenomena are under-represented in the database. Therefore, a model fitted on these large-scale natural databases will not reproduce these under-represented basic phenomena that could otherwise be easily illustrated with well selected artificial stimuli. In this work we study a specific example of the above statement. A standard cortical model using wavelets and divisive normalization tuned to reproduce subjective opinion on a large image quality dataset fails to reproduce basic cross-masking. Here we outline a solution for this problem by using artificial stimuli and by proposing a modification that makes the model easier to tune. Then, we show that the modified model is still competitive in the large-scale database. Our simulations with these artificial stimuli show that when using steerable wavelets, the conventional unit norm Gaussian kernels in divisive normalization should be multiplied by high-pass filters to reproduce basic trends in masking. Basic visual phenomena may be misrepresented in large natural image datasets but this can be solved with model-interpretable stimuli. This is an additional argument in line with Rust and Movshon (2005).

Maydis leaf blight (MLB) of maize ( Zea Mays L. ), a serious fungal disease, is capable of causing up to 70% damage to the crop under severe conditions. Severity of diseases is considered as one of the important factors for proper crop management and overall crop yield. Therefore, it is quite essential to identify the disease at the earliest possible stage to overcome the yield loss. In this study, we created an image database of maize crop, MDSD (Maydis leaf blight Disease Severity Dataset), containing 1,760 digital images of MLB disease, collected from different agricultural fields and categorized into four groups viz. healthy, low, medium and high severity stages. Next, we proposed a lightweight convolutional neural network (CNN) to identify the severity stages of MLB disease. The proposed network is a simple CNN framework augmented with two modified Inception modules, making it a lightweight and efficient multi-scale feature extractor. The proposed network reported approx. 99.13% classification accuracy with the f1-score of 98.97% on the test images of MDSD. Furthermore, the class-wise accuracy levels were 100% for healthy samples, 98% for low severity samples and 99% for the medium and high severity samples. In addition to that, our network significantly outperforms the popular pretrained models, viz. VGG16, VGG19, InceptionV3, ResNet50, Xception, MobileNetV2, DenseNet121 and NASNetMobile for the MDSD image database. The experimental findings revealed that our proposed lightweight network is excellent in identifying the images of severity stages of MLB disease despite complicated background conditions.

This study considers a method for reconstructing a high dynamic range (HDR) original image from a single saturated low dynamic range (LDR) image of metallic objects. A deep neural network approach was adopted for the direct mapping of an 8-bit LDR image to HDR. An HDR image database was first constructed using a large number of various metallic objects with different shapes. Each captured HDR image was clipped to create a set of 8-bit LDR images. All pairs of HDR and LDR images were used to train and test the network. Subsequently, a convolutional neural network (CNN) was designed in the form of a deep U-Net-like architecture. The network consisted of an encoder, a decoder, and a skip connection to maintain high image resolution. The CNN algorithm was constructed using the learning functions in MATLAB. The entire network consisted of 32 layers and 85,900 learnable parameters. The performance of the proposed method was examined in experiments using a test image set. The proposed method was also compared with other methods and confirmed to be significantly superior in terms of reconstruction accuracy, histogram fitting, and psychological evaluation.

Objectives Researchers have been studying the most important environmental cues that influence people’s fear and how to measure these emotions and perceptions in a more valid way. In order to contribute to experimental studies, we develop and validate the Urban Security Image Database (USID). Method The construction and validation of the USID followed two stages: (a) the obtainment by researchers of more than 300 naturalistic pictures in different urban contexts of the city of Porto and (b) using a within-subject design, a large sample ( N  = 1780) classified 49 selected pictures for fear of crime, risk perception of victimization, arousal, and valence levels. Results The validated Urban Security Image Database (USID) contains 49 pictures that are grouped in three categories according to fear mean levels: low fear, neutral fear, and high fear. Pictures of the low fear group depict residential areas, with high prospect spaces and well-cared vegetation. Pictures in the high fear group represent scenarios in night-time, with signs of incivilities and low prospect spaces. Fear of crime was negatively correlated with valence and positively with arousal. Conclusions USID is an important step to laboratorial experiments in the field of fear of crime and its relationship with environmental features. Moreover, since fear of crime is correlated with valence and arousal, it provides strength to the importance of considering fear a context-specific experience.

An image database of printed fabrics with repeating dot patterns was created to alleviate issues associated with management of and searches for numerous dot printed fabrics in the printing industry. The function of the database is to archive and allow retrieval of images. First, we discuss image archiving of repeating pattern-based dot printed fabrics. The color image was scanned by resolution of 200 dpi. The wavelet transformation was used to preprocess the image to obtain a scanned image 1/16 of the size of the original to be the stored image. To acquire images with repeating pattern color and repeating pattern template, the binary image of each pattern was obtained using the Sobel edge detection method and a morphological operation. Then pattern elements identical to the target pattern element were screened out. Afterwards, the centroid positions of these identical pattern elements were used to subdivide the repeating pattern color image and repeating pattern template image using a vertical vector method. Finally, the RGB 512-color histogram was used as the color feature of the dot printed fabrics, and the geometric and moment invariant feature values of the repeating pattern template image were used as the pattern feature of the dot printed fabrics. Our experimental results show that images can be acquired that are suitable for use in a dot printed fabric image database. The color and template images of the repeating patterns, which represent the image content of the printed fabrics, were obtained to create an image database of repeating pattern-based dot printed fabrics. This image database contains data on 300 printed fabrics which can be used for subsequent research on database image retrieval.

Cluster analysis is aimed at classifying elements into categories on the basis of their similarity. Its applications range from astronomy to bioinformatics, bibliometrics, and pattern recognition. We propose an approach based on the idea that cluster centers are characterized by a higher density than their neighbors and by a relatively large distance from points with higher densities. This idea forms the basis of a clustering procedure in which the number of clusters arises intuitively, outliers are automatically spotted and excluded from the analysis, and clusters are recognized regardless of their shape and of the dimensionality of the space in which they are embedded. We demonstrate the power of the algorithm on several test cases.

In this paper we propose a deep learning solution to age estimation from a single face image without the use of facial landmarks and introduce the IMDB-WIKI dataset, the largest public dataset of face images with age and gender labels. If the real age estimation research spans over decades, the study of apparent age estimation or the age as perceived by other humans from a face image is a recent endeavor. We tackle both tasks with our convolutional neural networks (CNNs) of VGG-16 architecture which are pre-trained on ImageNet for image classification. We pose the age estimation problem as a deep classification problem followed by a softmax expected value refinement. The key factors of our solution are: deep learned models from large data, robust face alignment, and expected value formulation for age regression. We validate our methods on standard benchmarks and achieve state-of-the-art results for both real and apparent age estimation.

Following the completion of the Great Wall Resource Survey in 2012, numerous landscape heritage resources along the Great Wall remained undiscovered, highlighting the limitations of conventional survey methods. This study aimed to conduct in-depth investigations of Great Wall signal sites through digital fieldwork methods, unveiling a crucial signaling structure—the auxiliary beacon—and presenting genuine historical scenes of the Great Wall signal network. Through the retrieval of the image database of the entire Great Wall and the utilization of UAVs (drones) for low-altitude remote sensing surveys, 252 auxiliary beacon sites were identified in diverse environments (e.g., deserts, mountains, plains) in Xinjiang, Gansu, Inner Mongolia, Qinghai, Ningxia, and other 10 regions. These case studies enable the categorization of layout types and the proposal of reconstruction hypotheses for the signal network of the Great Wall of China. The findings demonstrate that the beacon fire signals are not lit on the beacon tower tops, but through the ignition of various signals by auxiliary beacons, expressing pre-arranged information. Beacon towers and auxiliary beacons together form an efficient signal network along the Great Wall. This study explores how to use digital survey methods to unearth unknown landscape heritage resources of the Great Wall, enhancing the accuracy of observation for cross-regional and large-scale cultural heritage.