Hasil untuk "Photography"

Junning Xu, Sanxin Jiang

There are two problems with traditional knowledge distillation methods in industrial anomaly detection: first, traditional methods mostly use feature alignment between the same layers. The second is that similar or even identical structures are usually used to build teacher-student models, thus limiting the ability to represent anomalies in multiple ways. To address these issues, this work proposes a Hierarchical Knowledge Transfer (HKT) framework for detecting industrial surface anomalies. First, HKT utilizes the deep knowledge of the highest feature layer in the teacher’s network to guide student learning at every level, thus enabling cross-layer interactions. Multiple projectors are built inside the model to facilitate the teacher in transferring knowledge to each layer of the student. Second, the teacher-student structural symmetry is decoupled by embedding Convolutional Block Attention Modules (CBAM) in the student network. Finally, based on HKT, a more powerful anomaly detection model, HKT+, is developed. By adding two additional convolutional layers to the teacher and student networks of HKT, HKT+ achieves enhanced detection capabilities at the cost of a relatively small increase in model parameters. Experiments on the MVTec AD and BeanTech AD(BTAD) datasets show that HKT+ achieves state-of-the-art performance with average area under the receiver operating characteristic curve (AUROC) scores of 98.69% and 94.58%, respectively, which outperforms most current state-of-the-art methods.

Stanimir Stamatev, Miglena Angelova

The primary objective of the current research is to identify the impact that COVID-19 had on the Creative Business in Bulgaria and their first steps for recovery from the crisis. Our research is based on official and reliable Bulgarian National Statistical Insti-tute data. Based on the number of employed persons, the sectors most affected by COVID-19 in the creative business are sound recording and music publishing, photography, and advertising. For the period 2017-2022, there has been a slowdown in the growth rate of gross wages in the creative industries. This slowdown is most noticeable in 2019 and mainly in 2020, followed by a smooth and gradual recovery. At the same time, creative business companies in Bulgaria received the most extensive support in terms of public funding in 2020. The most generously supported sectors in 2020 within the Creative Industries are Advertising, followed by the Production and distribution of films and television shows and Computer Programming. The Creative Business companies in Bulgaria demonstrated relatively good sustainability and adaptability during and after the COVID-19 crisis. The rates in profit change confirm the serious negative impact of the pandemic (results in 2019 and 2020). Still, at the same time, it reveals a good level of resilience and recovery of the companies (results for 2021 and 2022).

Annosha Mubasher, Tim F. Liao

This research is concerned with individuals’ symbolic interactions with the Charging Bull statue. We chose this statue because it is an extremely popular tourist attraction and the most visited nonhuman statue in New York City. However, there is no research devoted to understanding how visitors interact with the statue. To fill this gap in the literature, we examine people’s interactions as a unique way of communicating with the Charging Bull statue in New York City through a photo analysis approach. Using the general framework of symbolic interactionism, the paper investigates poses, gestures, and more generally interactions of individuals near or with the statue to gain insights into the impact of the Charging Bull beyond a mere visual representation. The findings reveal various forms of interaction, including souvenir photography, active participation, unexpected interactions, and public-made symbolism. This study of these interactions contributes to the literature on collective memory and the study of statues. Additionally, this research highlights the inventiveness of the dynamic relationship between individuals and statues as symbols in the urban environment.

Alireza Saberironaghi, Jing Ren

Detecting cracks in the pavement is a vital component of ensuring road safety. Since manual identification of these cracks can be time-consuming, an automated method is needed to speed up this process. However, creating such a system is challenging due to factors including crack variability, variations in pavement materials, and the occurrence of miscellaneous objects and anomalies on the pavement. Motivated by the latest progress in deep learning applied to computer vision, we propose an effective U-Net-shaped model named DepthCrackNet. Our model employs the Double Convolution Encoder (DCE), composed of a sequence of convolution layers, for robust feature extraction while keeping parameters optimally efficient. We have incorporated the TriInput Multi-Head Spatial Attention (TMSA) module into our model; in this module, each head operates independently, capturing various spatial relationships and boosting the extraction of rich contextual information. Furthermore, DepthCrackNet employs the Spatial Depth Enhancer (SDE) module, specifically designed to augment the feature extraction capabilities of our segmentation model. The performance of the DepthCrackNet was evaluated on two public crack datasets: Crack500 and DeepCrack. In our experimental studies, the network achieved mIoU scores of 77.0% and 83.9% with the Crack500 and DeepCrack datasets, respectively.

Mabel K. Yan, Anne E. Cust, H. Peter Soyer et al.

Abstract Introduction Melanoma surveillance photography (MSP) is a comprehensive surveillance method that comprises two- or three-dimensional total body photography with tagged digital dermoscopy, performed at prescribed intervals. It has the potential to reduce unnecessary biopsies and enhance early detection of melanoma, but it is not yet standard care for all high-risk patients in Australia. This protocol describes a randomised controlled trial (RCT) designed to evaluate the clinical impact and cost-effectiveness of using MSP for the surveillance of individuals at ultra-high or high risk of melanoma from a health system perspective. Methods and design This is a registry-based, unblinded, multi-site, parallel-arm RCT that will be conducted over 3 years. We aim to recruit 580 participants from three Australian states: Victoria, New South Wales and Queensland, via state cancer registries or direct referral from clinicians. Eligible participants within 24 months of a primary cutaneous melanoma diagnosis will be randomised 1:1 to receive either MSP in addition to their routine clinical surveillance (intervention group) or routine clinical surveillance without MSP (control group). Most participants will continue surveillance with their usual care provider, and the frequency of follow-up visits in both groups will depend on the stage of their primary melanoma and risk factors. The primary outcome measure of the study is the number of unnecessary biopsies (i.e. false positives, being cases where a lesion is biopsied due to suspected melanoma on clinical examination, either with or without MSP, but the resulting histopathology finding is negative for melanoma). Secondary outcomes include the evaluation of health economic outcomes, quality of life and patient acceptability. Two sub-studies will explore the benefit of MSP in high-risk patients prior to a melanoma diagnosis and the diagnostic performance of MSP in the teledermatology setting compared to the en face clinical setting. Discussion This trial will determine the clinical efficacy, cost-effectiveness and affordability of MSP to facilitate policy decision-making at the national and local levels, across primary and specialist care. Trial registration ClinicalTrials.gov NCT04385732 . Registered on May 13, 2020.

Linlong Jing, Xinhua Wei, Qi Song et al.

Open-canopy pesticide application technology is not affected by the canopy structure and can substantially increase droplet penetration and deposition in the middle and lower rice canopy. However, the optimal working position of the canopy opener is difficult to determine. The purpose of this study was to examine the characteristics of rice canopy disturbances under the action of a canopy opener, verify simulation results using high-speed photography tests, and identify the spraying effect of the optimal operating parameters of a canopy opener through spray tests. The mechanical characteristic parameters of the rice heading stage were determined through experiments, and the canopy-opening process was simulated based on the finite element method of explicit dynamics. The simulation results show that when the speed of the canopy opener is 1.2 m s⁻¹ and the height is 0.6 m, the rice canopy is the most disturbed. The spray experiment results show that under this condition, the overall coverage rate of the rice canopy is relatively high, and the coverage rates of the upper, middle, and lower layers of rice are 92.18%, 66.43%, and 52.84%, respectively.

Yuanyuan Liu, Ming He, Yueyong Wang et al.

The main black land conservation measure in China is the straw return to the fields. The processing of high-resolution images collected by aerial photography of UAVs through image stitching technology can provide image information for achieving fast and accurate detection of straw cover over large areas. The classical SIFT algorithm has many drawbacks, such as high dimensionality of feature descriptors, high computational effort, and low matching efficiency. To solve the problems above, this study proposes an improved algorithm. First, the method down sampled the high-resolution images before detecting the features to reduce the number of feature points and improve the efficiency of feature detection. Then, matching among feature points is achieved by gradient normalization-based feature descriptors to improve the matching accuracy. Next, the Progressive Sample Consistency algorithm eliminates the mismatch points and optimizes the transformation model. Finally, the images are fused with optimal stitching combined with fade-in and fade-out to achieve high-quality stitching. The comparative experimental results show that compared with the traditional SIFT and the speed-up robust feature algorithms, the algorithm has the advantage of the speed and good robustness to angle rotation, and makes full use of the texture information and the detail information, so it has higher accuracy. Compared with the traditional methods, the panoramic stitching image quality herein is excellent and can be applied to subsequent straw cover detection, the straw cover error is ≤3%, meeting the demand for large-area straw cover detection. Overall, the method proposed herein achieves an ideal balance between accuracy and efficiency; and outperforms other widely used and superior methods.

Daniel Queirós da Silva, Filipe Neves dos Santos, Armando Jorge Sousa et al.

Mobile robotics in forests is currently a hugely important topic due to the recurring appearance of forest wildfires. Thus, in-site management of forest inventory and biomass is required. To tackle this issue, this work presents a study on detection at the ground level of forest tree trunks in visible and thermal images using deep learning-based object detection methods. For this purpose, a forestry dataset composed of 2895 images was built and made publicly available. Using this dataset, five models were trained and benchmarked to detect the tree trunks. The selected models were SSD MobileNetV2, SSD Inception-v2, SSD ResNet50, SSDLite MobileDet and YOLOv4 Tiny. Promising results were obtained; for instance, YOLOv4 Tiny was the best model that achieved the highest AP (90%) and F1 score (89%). The inference time was also evaluated, for these models, on CPU and GPU. The results showed that YOLOv4 Tiny was the fastest detector running on GPU (8 ms). This work will enhance the development of vision perception systems for smarter forestry robots.

Xin Zeng, Qiang Guo, Haoran Duan et al.

Abstract Crowd counting is still a practical and challenging problem owing to scale variations and information loss. Most existing methods based on the straightforward fusion of different features from a deep neural network seem to eliminate this limitation. However, these features are difficult to be fused since they often differ significantly in modality and dimensionality. Unlike previous works, a multi‐level features extraction network with gating mechanism for crowd counting is proposed. Specifically, a multi‐channel gated unit to adaptively extract features in different levels of the network is proposed, which can avoid interference from confusing information. To fully aggregate features via multi‐level fusion, multi‐level features extraction scheme is presented. The multi‐level features extraction network learns to fuse features from multiple levels and reduce false predictions. Extensive experiments and evaluations clearly illustrate that the proposed approach achieves state‐of‐the‐art counting performance against other methods on four mainstream crowd counting benchmarks.

Patricio Martínez-Carricondo, Francisco Agüera-Vega, Fernando Carvajal-Ramírez

In this study, an analysis of the capabilities of unmanned aerial vehicle (UAV) photogrammetry to obtain point clouds from areas with a near-vertical inclination was carried out. For this purpose, 18 different combinations were proposed, varying the number of ground control points (GCPs), the adequacy (or not) of the distribution of GCPs, and the orientation of the photographs (nadir and oblique). The results have shown that under certain conditions, the accuracy achieved was similar to those obtained by a terrestrial laser scanner (TLS). For this reason, it is necessary to increase the number of GCPs as much as possible in order to cover a whole study area. In the event that this is not possible, the inclusion of oblique photography ostensibly improves results; therefore, it is always advisable since they also improve the geometric descriptions of break lines or sudden changes in slope. In this sense, UAVs seem to be a more economic substitute compared to TLS for vertical wall surveying.

Niels Anders, João Valente, Rens Masselink et al.

Digital Elevation Models (DEMs) are 3D representations of the Earth’s surface and have numerous applications in geomorphology, hydrology and ecology. Structure-from-Motion (SfM) photogrammetry using photographs obtained by unmanned aerial vehicles (UAVs) have been increasingly used for obtaining high resolution DEMs. These DEMs are interpolated from point clouds representing entire landscapes, including points of terrain, vegetation and infrastructure. Up to date, there has not been any study clearly comparing different algorithms for filtering of vegetation. The objective in this study was, therefore, to assess the performance of various vegetation filter algorithms for SfM-obtained point clouds. The comparison was done for a Mediterranean area in Murcia, Spain with heterogeneous vegetation cover. The filter methods that were compared were: color-based filtering using an excessive greenness vegetation index (VI), Triangulated Irregular Networks (TIN) densification from LAStools, the standard method in Agisoft Photoscan (PS), iterative surface lowering (ISL), and a combination of iterative surface lowering and the VI method (ISL_VI). Results showed that for bare areas there was little to no difference between the filtering methods, which is to be expected because there is little to no vegetation present to filter. For areas with shrubs and trees, the ISL_VI and TIN method performed best. These results show that different filtering techniques have various degrees of success in different use cases. A default filter in commercial software such as Photoscan may not always be the best way to remove unwanted vegetation from a point cloud, but instead alternative methods such as a TIN densification algorithm should be used to obtain a vegetation-less Digital Terrain Model (DTM).

Fousia M. Shamsudeen, G. Raju

Fundus photography is a powerful imaging modality that is utilized for detecting macular degeneration, retinal neoplasms, choroid disturbances, glaucoma and diabetic retinopathy. As the illumination source in fundus imaging is situated at the center of the fundus camera, the illumination at the peripheral regions of the images would be relatively less than the center, which is termed vignetting. Vignetting adversely affects the performance of computerized methods for analyzing fundus imagery. A devignetting method for fundus imagery based on the Modified Sigmoid Transform (MST) is proposed in this paper. Gain (A) and centering parameter (α) of MST have a crucial influence on its performance. For low values of the gain, local contrast is penalized, and the overall dynamic range is compressed. When the value of gain is very high, the images after the illumination correction will have a washed out appearance. The optimum value of gain is determined in this paper from an objective method based on two statistical indices, Average Gradient of Illumination Component (AGIC) and Error of Enhancement (EME). MST with gain value defined via objective methods is able to correct the uneven illumination in fundus images without penalizing the local contrast. The proposed method is compared with illumination equalization model, homomorphic filtering and Adaptive Gamma Correction (AGC) and was found to be superior in terms of ‘naturality’, uniformity of background illumination, and computational speed. Keywords: Average gradient of illumination component, Devignetting, Error of enhancement, Fundus imagery, Illumination correction, Modified sigmoid transform (MST), Objective function

A. Audi, A. Audi, M. Pierrot-Deseilligny et al.

In the recent years, unmanned aerial vehicles (UAVs) have become an interesting tool in aerial photography and photogrammetry activities. In this context, some applications (like cloudy sky surveys, narrow-spectral imagery and night-vision imagery) need a longexposure time where one of the main problems is the motion blur caused by the erratic camera movements during image acquisition. This paper describes an automatic real-time stacking algorithm which produces a high photogrammetric quality final composite image with an equivalent long-exposure time using several images acquired with short-exposure times. <br><br> Our method is inspired by feature-based image registration technique. The algorithm is implemented on the light-weight IGN camera, which has an IMU sensor and a SoC/FPGA. To obtain the correct parameters for the resampling of images, the presented method accurately estimates the geometrical relation between the first and the <i>N^th</i> image, taking into account the internal parameters and the distortion of the camera. Features are detected in the first image by the FAST detector, than homologous points on other images are obtained by template matching aided by the IMU sensors. The SoC/FPGA in the camera is used to speed up time-consuming parts of the algorithm such as features detection and images resampling in order to achieve a real-time performance as we want to write only the resulting final image to save bandwidth on the storage device. The paper includes a detailed description of the implemented algorithm, resource usage summary, resulting processing time, resulting images, as well as block diagrams of the described architecture. The resulting stacked image obtained on real surveys doesn’t seem visually impaired. Timing results demonstrate that our algorithm can be used in real-time since its processing time is less than the writing time of an image in the storage device. An interesting by-product of this algorithm is the 3D rotation estimated by a photogrammetric method between poses, which can be used to recalibrate in real-time the gyrometers of the IMU.

Pritam Bawankar, Nita Shanbhag, S Smitha K et al.

Diabetic retinopathy (DR) is a leading cause of blindness among working-age adults. Early diagnosis through effective screening programs is likely to improve vision outcomes. The ETDRS seven-standard-field 35-mm stereoscopic color retinal imaging (ETDRS) of the dilated eye is elaborate and requires mydriasis, and is unsuitable for screening. We evaluated an image analysis application for the automated diagnosis of DR from non-mydriatic single-field images. Patients suffering from diabetes for at least 5 years were included if they were 18 years or older. Patients already diagnosed with DR were excluded. Physiologic mydriasis was achieved by placing the subjects in a dark room. Images were captured using a Bosch Mobile Eye Care fundus camera. The images were analyzed by the Retinal Imaging Bosch DR Algorithm for the diagnosis of DR. All subjects also subsequently underwent pharmacological mydriasis and ETDRS imaging. Non-mydriatic and mydriatic images were read by ophthalmologists. The ETDRS readings were used as the gold standard for calculating the sensitivity and specificity for the software. 564 consecutive subjects (1128 eyes) were recruited from six centers in India. Each subject was evaluated at a single outpatient visit. Forty-four of 1128 images (3.9%) could not be read by the algorithm, and were categorized as inconclusive. In four subjects, neither eye provided an acceptable image: these four subjects were excluded from the analysis. This left 560 subjects for analysis (1084 eyes). The algorithm correctly diagnosed 531 of 560 cases. The sensitivity, specificity, and positive and negative predictive values were 91%, 97%, 94%, and 95% respectively. The Bosch DR Algorithm shows favorable sensitivity and specificity in diagnosing DR from non-mydriatic images, and can greatly simplify screening for DR. This also has major implications for telemedicine in the use of screening for retinopathy in patients with diabetes mellitus.

Grzegorz Szewczyk, Dariusz Kulak, Arkadiusz Stańczykiewicz et al.

The study covered an analysis of the accuracy level of measuring time within a working shift using the method of regular snapshot observation at a harvester operator’s worksite in Scots pine stands. A conformance level of the analyzed methods was evaluated through assessing the accuracy of rectilinear fitting of time structures, established using the photography of work day method and snapshot observations. The accuracy of snapshot measurements performed in 3-minute intervals was determined as high, exceeding 95%. Increasing the time interval between observations to 10 or 15 minutes resulted in higher estimation error in snapshot observation time, ranging between 5 and 10% for late thinned and clear-cut stands. The accuracy of evaluating proportions of specific work times within a working shift, in regular snapshot observations, was correlated with work cyclicality. The strongest work cycle in thinned stands consisted of 43 activities, with total duration of 13 minutes, whereas in clearcut stands it comprised 45 activities, with total duration of 15 minutes. One of the advantages of the described method, apart from its lower labour intensity as compared to working day photography, was the possibility to assess labour time and breaks as well as estimate the share of downtime.

Wei Chen, Ying Wu, Mi Zheng et al.

<b>AIM:</b>To develop a reliable, reproducible rat model of retinal vein occlusion (RVO) with a novel photosensitizer (erythrosin B) and study the cellular responses in the retina.<b>METHODS</b>:Central and branch RVOs were created in adult male rats <i>via</i> photochemically-induced ischemia. Retinal changes were monitored<i> via</i> color fundus photography and fluorescein angiography at 1 and 3h, and 1, 4, 7, 14, and 21d after irradiation. Tissue slices were evaluated histopathologically. Retinal ganglion cell survival at different times after RVO induction was quantified by nuclear density count. Retinal thickness was also observed.<b>RESULTS</b>:For all rats in both the central and branch RVO groups, blood flow ceased immediately after laser irradiation and retinal edema was evident at one hour. The retinal detachment rate was 100% at 3h and developed into bullous retinal detachment within 24h. Retinal hemorrhages were not observed until 24h. Clearance of the occluded veins at 7d was observed by fluorescein angiography. Disease manifestation in the central RVO eyes was more severe than in the branch RVO group. A remarkable reduction in the ganglion cell count and retinal thickness was observed in the central RVO group by 21d, whereas moderate changes occurred in the branch RVO group.<b>CONCLUSION:</b> Rat RVO created by photochemically-induced ischemia using erythrosin B is a reproducible and reliable animal model for mimicking the key features of human RVO. However, considering the 100% rate of retinal detachment, this animal model is more suitable for studying RVO with chronic retinal detachment.

Aleksandra Hirszfeld

Labor in a single shot: a history of 21st-century professional activities through a camera lens The text deals with the topic of an international series of workshops initiated by Antje Ehman and Harun Farocki in 2011 under the auspices of the Goethe Institute and dedicated to the issue of labour. At present, ten workshops have been conducted of a planned fifteen (including Hanoi, Buenos Aires, Rio de Janeiro, Cairo, Berlin, Tel Aviv, Geneva and Lisbon). Łódź, representing Poland, was the ninth-place city. The Polish version of the workshop was attended by visual artists from the National Film School. The results of the workshop will be exhibited at the October exposition in 2013 and the Łódź Museum of Art. The curators of the project are Aleksandra Jach and Joanna Sokołowska

Pravin Salunke, Amey Savardekar, Sukumar Sura

Bilateral sixth nerve paresis following closed head injury, though rare, is a known entity. However, delayed-onset post-traumatic bilateral abducens paresis is extremely rare. We present two cases. The first patient had onset of bilateral abducens paresis 2 weeks after closed head injury and the second patient after 3 days. The cause in the former was detected to be chronic subdural hematoma and in the latter is speculated to be edema/ischemia due to injury to soft tissue structures housing these nerves. The delayed onset of bilateral abducens paresis following head injury may vary according to the cause. There may be another mechanism of injury apart from direct trauma. Though rare, it needs to be evaluated and may have a treatable cause like elevated intracranial pressure.

Margaret R. Higonnet

During World War I not only did technical revolutions produce a new, devastatingly “total” type of war, but they also changed the way war was seen, both from a medical and a military perspective. While the medical gaze probed the interior of bodies, the military gaze pulled away from the immediacy of the combatant’s gaze. The accompanying gender revolution placed women in new positions, both on the front line and behind the camera. Using the work of professional photographer Olive Edis this article first explores the three testimonial functions of photography concerning the photographer’s identity, tasks and physical location, before turning to the memoirs of two Red Cross workers, Florence Farmborough and Margaret Hall, which suggest that a particular war genre arose at the moment when women created scrapbooks that combined photographs with their memoirs: the “photo-text”.

Carlos Ruiz, Andrés Trujillo, Alvin García

El estándar DICOM (Digital Imaging and Communication in Medicine) es un protocolo no propietario para el intercambio de información médica. DICOM representa y define la información de objetos del mundo real tales como una resonancia magnética (MRI), una tomografía computarizada (CT) y una fotografía médica digital (VL Photographic), por medio de definiciones de objeto de información llamados IOD. El presente artículo describe una metodología para representar el IOD de una fotografía médica digital de luz visible (VL Photographic Image) por intermedio de documentos XML Schema. Estos documentos se utilizan en la creación y validación de documentos XML para representar información clínica técnica asociada a fotografías médicas digitales para su posterior implementación en una aplicación web de teledermatología.<br>DICOM standard (Digital Imaging and Communication in Medicine) is a non-proprietary protocol for the medical exchange information. DICOM represents and defines the information of real world objects like a magnetic resonance image (MRI), a computerized axial tomography and a digital medical photography (VL photographic), through information object definitions called IOD. The present article describes a methodology to represent the IOD of a digital medical photography of visible light (VL Photographic Image) through XML Schema documents. These documents were used in the creation and validation of XML documents to represent digital medical photographies compiling clinical and technical information for their later implementation in a teledermatology application.