Sharpening Filter Research Articles

A Comprehensive Study on Image Segmentation Using Pulse-Coupled Neural Networks

Pulse-Coupled Neural Network (PCNN) operates as a matrix of neurons, each uniquely corresponding to a pixel in the image being processed. Unlike traditional neural networks, PCNN does not require any training, making it highly suitable for image segmentation tasks. However, the complexity of PCNN arises from its numerous parameters, making parameter selection a challenging endeavor. In this work, we introduce a simplified PCNN architecture that includes an automatic parameter determination method. This approach is specifically designed for binary image segmentation and has been tested on various images, yielding results with distinct and desirable features. The proposed network iterates only four times, enhancing its efficiency. During pre-processing, RGB images are converted to HSV color space, and the V component undergoes further processing. This component is first filtered using an averaging filter, followed by a sharpening filter. The parameters for the PCNN are then generated automatically, eliminating the need for manual selection and making the network highly suitable for real-time image processing. The performance of the proposed network has been verified through tests on a variety of images. Key Words: Neural Networks, Pulse Coupled Neural Networks, Image Processing

Vertical root fracture diagnosis in teeth with metallic posts: Impact of metal artifact reduction and sharpening filters.

This study examined the influence of a metal artifact reduction (MAR) tool, sharpening filters, and their combination on the diagnosis of vertical root fracture (VRF) in teeth with metallic posts using cone-beam computed tomography (CBCT). Twenty single-rooted human premolars - 9 with VRF and 11 without - were individually placed in a human mandible. A metallic post composed of a cobalt-chromium alloy was inserted into the root canal of each tooth. CBCT scans were then acquired under the following parameters: 8 mA, a 5×5 cm field of view, a voxel size of 0.085 mm, 90 kVp, and with MAR either enabled or disabled. Five oral and maxillofacial radiologists independently evaluated the CBCT exams under each MAR mode and across 3 sharpening filter conditions: no filter, Sharpen 1×, and Sharpen 2×. The diagnostic performance was quantified by the area under the receiver operating characteristic curve (AUC), sensitivity, and specificity. These metrics were compared using 2-way analysis of variance with a significance level of α=5%. Intra- and inter-examiner agreement were assessed using the weighted kappa test. Neither MAR nor the application of sharpening filters significantly impacted AUC or specificity (P>0.05). However, sensitivity increased when MAR was combined with Sharpen 1× and Sharpen 2× (P=0.015). The intra-examiner agreement ranged from fair to substantial (0.34-0.66), while the inter-examiner agreement ranged from fair to moderate (0.27-0.41). MAR in conjunction with sharpening filters improved VRF detection; therefore, their combined use is recommended in cases of suspected VRF.

Combination of metal artifact reduction and sharpening filter application for horizontal root fracture diagnosis in teeth adjacent to a zirconia implant.

This study examined the influence of metal artifact reduction (MAR), the application of sharpening filters, and their combination on the diagnosis of horizontal root fracture (HRF) in teeth adjacent to a zirconia implant on cone-beam computed tomography (CBCT) examinations. Nineteen single-rooted teeth (9 with HRF and 10 without) were individually positioned in the right central incisor socket of a dry human maxilla. A zirconia implant was placed adjacent to each tooth. Imaging was performed using an OP300 Maxio CBCT (Instrumentarium, Tuusula, Finland) unit with the following settings: a current of 8 mA, both MAR modes (enabled and disabled), a 5×5 cm field of view, a voxel size of 0.085 mm, and a peak kilovoltage of 90 kVp. Four oral and maxillofacial radiologists independently evaluated the CBCT scans under both MAR conditions and across 3 levels of sharpening filter application (none, Sharpen 1×, and Sharpen 2×). Diagnostic metrics were calculated and compared using 2-way analysis of variance (α=5%). The weighted kappa test was used to assess intra- and inter-examiner reliability in the diagnosis of HRF. MAR tool activation, sharpening filter use, and their combination did not significantly impact the area under the receiver operating characteristic curve, sensitivity, or specificity of HRF diagnosis (P>0.05). Intra- and inter-examiner agreement ranged from fair to substantial. The diagnosis of HRF in a tooth adjacent to a zirconia implant is not affected by the activation of MAR, the application of a sharpening filter, or the combination of these tools.

The spectral stacking method and its application in seismic data resolution increase

AbstractPreconditioning is a practice adopted in the treatment of seismic data and, when properly applied, favours a resolution increase. However, retrieving all the frequency contents is still a challenge to overcome. Techniques to improve the seismic resolution should preserve the existing relations in the distribution of seismic events. Otherwise, it may cause undesirable situations, from the degradation of the signal‐to‐noise ratio to the creation of false events, leading to errors of interpretation. We developed a method called spectral stacking, which makes use of circular convolution in the frequency domain to extend frequency content, preserving the distribution of seismic events without adding artefacts. As an application of the spectral stacking method, we create a sharpening filter that resolves layers under Ricker's criterion. In addition, the sharpening filter preserves the amplitude using a variable‐window correction based on the local minima of the absolute values of amplitude. We validated our proposed method in synthetic and real tests. The results showed that the sharpening filter obtained a resolution increase, providing a sharper image of the seismic section.

Pengaruh Data Preprocessing terhadap Imbalanced Dataset pada Klasifikasi Citra Sampah menggunakan Algoritma Convolutional Neural Network

Garbage is one of Indonesia's most significant problems with an increase in waste each year reaching 187.2 million tonnes/year. Various efforts to reduce the amount of waste such as Garbage Banks have been encouraged. However, this program has not run well, because some people have difficulty distinguishing the type of waste. One solution to overcome this problem is that need a system that can classify the type of waste. The deep learning approach with the CNN algorithm is currently widely used to solve classification problems. This method requires a large number of datasets to increase the level of accuracy. Getting a garbage dataset is a particular problem in the training process because the dataset is unbalanced. The dataset used amounted to 2527 data consisting of 6 classes. Several treatments such as undersampling and image augmentation are applied to overcome imbalanced datasets. Other treatments such as the type of input image channel and the use of filters are combined into 24 experimental scenarios to achieve the highest accuracy. The results of the experiment get the best scenario, namely, the dataset is undersampling and then augmented with 5 geometric transformation parameters with the input image being RGB and applying a sharpening filter to get an accuracy value of 0.9919 with 20 epochs.

SCU-Net++: A Nested U-Net Based on Sharpening Filter and Channel Attention Mechanism

U-Net++ is one of the most prominent deep convolutional neural networks in the field of medical image segmentation after U-Net. However, the semantic gaps between the encoder and decoder subnets are still large, which will lead to fuzzy feature maps and even target regions of segmentation. To solve this problem, we propose an improved semantic segmentation model utilizing channel attention mechanism and Laplacian sharpening filter, SCU-Net++: dense skip connections are redesigned with sharpening filters to ease the semantic gaps, and channel attention modules are used to make the model pay more attention on the feature maps that are useful for our pixel-level classification task. Compared with U-Net++, the proposed model obtains a more competitive performance on the Pancreas Segmentation dataset and Liver Tumor Segmentation dataset, while increases a very small number of learnable parameters and thus almost does not make additional training and reasoning costs. The training of the proposed method is carried out in deep supervision mode, which alleviates the problem of gradient disappearance, and pruning mechanism can be activated to accelerate the reasoning speed.

Efficient 2D image Upscaling using Iterative Optimized Sharpening filtering

In real-time image interpolation, polynomial-based approaches are used because they are simple and practical. The unknown pixels in the image grid are obtained using polynomial-based upscaling by taking into account the weighted average of adjacent pixels. It causes degradation in high frequency (HF) parts of the image due to blurring artifacts. Numerous edge-directed and learning-based techniques are discussed in the literature, which also introduces blurring in the image’s high variance region. To address the aforementioned issue, one pre-processing method based on the concept of unsharp masking is proposed in order to generate sharpened high resolution (HR) images from low resolution (LR) images. The LR image is blurred using a weighted average filter, which is based on the concept of unsharp masking. The difference between the LR image and the blurred LR image is then extracted to represent the loss edge portion of the image. This error or degraded HF image is sharpened iteratively using a cuckoo optimized sharpening filter known as the Iteratively Optimized Sharpening (IOS) filter. The subtle is then collected by iteratively applying ISO and combining it with the LR image using one optimized gain factor, resulting in a sharpened LR image. This pre-processing is done prior to interpolation to compensate for the loss of HF details. To up-sample the sharpened LR image, the MAKIMA scheme is used. Because of MAKIMA, the image’s edges and boundaries are preserved. As a result, an HR image with sharpened edges and texture details can be obtained. In comparison to existing techniques, the proposed algorithm yields better results in terms of visual quality and objectivity. Various image databases are used to assess the efficiency of the proposed method.

Impact of Deep Learning Reconstruction Combined With a Sharpening Filter on Single-Shot Fast Spin-Echo T2-Weighted Magnetic Resonance Imaging of the Uterus.

This study aimed to evaluate the effects of deep learning (DL) reconstruction and a postprocessing sharpening filter on the image quality of single-shot fast spin-echo (SSFSE) T2-weighted imaging (T2WI) of the uterus. Fifty consecutive patients who underwent pelvic magnetic resonance imaging were included. Parasagittal T2WI with a slice thickness of 4 mm was obtained with the periodically rotated overlapping parallel lines with enhanced reconstruction (PROPELLER) and SSFSE sequences (mean scan time, 204 and 22 seconds, respectively). The following 3 types of SSFSE images were reconstructed, and the signal-to-noise ratio (SNR) and tissue contrast were assessed: conventional reconstruction (SSFSE-C), DL reconstruction (SSFSE-DL), and DL with a sharpening filter (SSFSE-DLF). Three radiologists independently assessed image quality, and area under the visual grading characteristics curve (AUCVGC) analysis was performed to compare the SSFSE and PROPELLER images. Compared with that of the PROPELLER images, the SNR of the SSFSE-C, SSFSE-DL, and SSFSE-DLF images was significantly lower (P < 0.05), significantly higher (P < 0.05), and equivalent, respectively. The SSFSE-DL images exhibited significantly lower contrast between the junctional zone and myometrium than those obtained with the other sequences (P < 0.05). In qualitative comparisons with the PROPELLER images, all 3 SSFSE sequences, SSFSE-DL, and SSFSE-DLF demonstrated significantly higher scores for artifacts, noise, and sharpness, respectively (P < 0.01). The overall image quality of SSFSE-C (mean AUCVGC, 0.03; P < 0.01) and SSFSE-DL (mean AUCVGC, 0.23; P < 0.01) was rated as significantly inferior, whereas that of SSFSE-DLF (mean AUCVGC, 0.69) was equivalent or significantly higher (P < 0.01). Using a combination of DL and a sharpening filter markedly increases the image quality of SSFSE of the uterus to the level of the PROPELLER sequence.

SSVEP-EEG Feature Enhancement Method Using an Image Sharpening Filter.

Steady-state visual evoked potential (SSVEP) is widely used in brain computer interface (BCI), medical detection, and neuroscience, so there is significant interest in enhancing SSVEP features via signal processing for better performance. In this study, an image processing method was combined with brain signal analysis and a sharpening filter was used to extract image details and features for the enhancement of SSVEP features. The results demonstrated that sharpening filter could eliminate the SSVEP signal trend term and suppress its low-frequency component. Meanwhile, sharpening filter effectively enhanced the signal-to-noise ratios (SNRs) of the single-channel and multi-channel fused signals. Image sharpening filter also significantly improved the recognition accuracy of canonical correlation analysis (CCA), filter bank canonical correlation analysis (FBCCA), and task-related component analysis (TRCA). The tools developed here effectively enhanced the SSVEP signal features, suggesting that image processing methods can be considered for improved brain signal analysis.

Detection of Surface Defects on Railway Tracks Based on Deep Learning

The detection of rail surface defects is very important in railway transportation. However, the edge defects on both sides of the rail and the multi-scale variation between different types of defects both pose challenges to the detection of rail surface defects. In order to solve the above problems, this paper proposes a novel rail surface defect detection network, YOLOv5s-VF. First, we design a sharpening functional attention mechanism (V-CBAM) that contains two key components: adaptive channel attention (F-CAM) and sharpened spatial attention (SSA). In F-CAM, we use one-dimensional convolution with adaptive convolution kernels for cross-channel connections, which reduces the number of parameters of the attention mechanism without affecting its performance. In SSA, we design a sharpening filter suitable for spatial attention, which is used to enhance the attention to the edge position defects of railway tracks and enhance the detection effect of the network on edge defects. Second, we construct a microscale adaptive spatial feature fusion (M-ASFF), which adds a high-resolution feature extraction layer to enhance the details of the underlying features of tiny defects. At the same time, in order to prevent the loss of detailed information and the excessive increase of the parameters of the model, the low-resolution feature layer is removed. Combined with adaptive spatial feature fusion, it can prevent the semantic conflict caused by the fusion of features at different scales. Finally, given the lack of labeled public rail surface defect datasets, this paper is based on the collection of real rail images and manually labels defects to train an object detection network and open source it. The experimental results show that YOLOv5s-VF outperforms the existing rail surface defect detection methods with a detection accuracy of 93.5% and a detection speed of 114.9 fps.

One Spatio-Temporal Sharpening Attention Mechanism for Light-Weight YOLO Models Based on Sharpening Spatial Attention.

Attention mechanisms have demonstrated great potential in improving the performance of deep convolutional neural networks (CNNs). However, many existing methods dedicate to developing channel or spatial attention modules for CNNs with lots of parameters, and complex attention modules inevitably affect the performance of CNNs. During our experiments of embedding Convolutional Block Attention Module (CBAM) in light-weight model YOLOv5s, CBAM does influence the speed and increase model complexity while reduce the average precision, but Squeeze-and-Excitation (SE) has a positive impact in the model as part of CBAM. To replace the spatial attention module in CBAM and offer a suitable scheme of channel and spatial attention modules, this paper proposes one Spatio-temporal Sharpening Attention Mechanism (SSAM), which sequentially infers intermediate maps along channel attention module and Sharpening Spatial Attention (SSA) module. By introducing sharpening filter in spatial attention module, we propose SSA module with low complexity. We try to find a scheme to combine our SSA module with SE module or Efficient Channel Attention (ECA) module and show best improvement in models such as YOLOv5s and YOLOv3-tiny. Therefore, we perform various replacement experiments and offer one best scheme that is to embed channel attention modules in backbone and neck of the model and integrate SSAM into YOLO head. We verify the positive effect of our SSAM on two general object detection datasets VOC2012 and MS COCO2017. One for obtaining a suitable scheme and the other for proving the versatility of our method in complex scenes. Experimental results on the two datasets show obvious promotion in terms of average precision and detection performance, which demonstrates the usefulness of our SSAM in light-weight YOLO models. Furthermore, visualization results also show the advantage of enhancing positioning ability with our SSAM.

A deep learning approach using effective preprocessing techniques to detect COVID-19 from chest CT-scan and X-ray images

Coronavirus disease-19 (COVID-19) is a severe respiratory viral disease first reported in late 2019 that has spread worldwide. Although some wealthy countries have made significant progress in detecting and containing this disease, most underdeveloped countries are still struggling to identify COVID-19 cases in large populations. With the rising number of COVID-19 cases, there are often insufficient COVID-19 diagnostic kits and related resources in such countries. However, other basic diagnostic resources often do exist, which motivated us to develop Deep Learning models to assist clinicians and radiologists to provide prompt diagnostic support to the patients. In this study, we have developed a deep learning-based COVID-19 case detection model trained with a dataset consisting of chest CT scans and X-ray images. A modified ResNet50V2 architecture was employed as deep learning architecture in the proposed model. The dataset utilized to train the model was collected from various publicly available sources and included four class labels: confirmed COVID-19, normal controls and confirmed viral and bacterial pneumonia cases. The aggregated dataset was preprocessed through a sharpening filter before feeding the dataset into the proposed model. This model attained an accuracy of 96.452% for four-class cases (COVID-19/Normal/Bacterial pneumonia/Viral pneumonia), 97.242% for three-class cases (COVID-19/Normal/Bacterial pneumonia) and 98.954% for two-class cases (COVID-19/Viral pneumonia) using chest X-ray images. The model acquired a comprehensive accuracy of 99.012% for three-class cases (COVID-19/Normal/Community-acquired pneumonia) and 99.99% for two-class cases (Normal/COVID-19) using CT-scan images of the chest. This high accuracy presents a new and potentially important resource to enable radiologists to identify and rapidly diagnose COVID-19 cases with only basic but widely available equipment.

Image Focus Enhancement Using Focusing Filter and DT-CWT Based Image Fusion

Combining multi-model images of the same scene that have different focus distances can produce clearer and sharper images with a larger depth of field. Most available image fusion algorithms are superior in results. However, they did not take into account the focus of the image. In this paper a fusion method is proposed to increase the focus of the fused image and to achieve highest quality image using the suggested focusing filter and Dual Tree-Complex Wavelet Transform. The focusing filter consist of a combination of two filters, which are Wiener filter and a sharpening filter. This filter is used before the fusion operation using Dual Tree-Complex Wavelet Transform. The common fusion rules, which are the average-fusion rule and maximum-fusion rule, were used to obtain the fused image. In the experiment, using the focus operators, the performance of the proposed fusion algorithm was compared with the existing algorithms. The results showed that the proposed method is better than these fusion methods in terms of the focus and quality.

A DWT Based Novel Multimodal Image Fusion Method

Nowadays multimodal image fusion has been majorly utilized as an important processing tool in various image related applications. For capturing useful information different sensors have been developed. Mainly such sensors are infrared (IR) image sensor and visible (VI) image sensor. Fusing both these sensors provides better and accurate scene information. The major application areas where this fused image has been mostly used are military, surveillance, and remote sensing. For better identification of targets and to understand overall scene information, the fused image has to provide better contrast and more edge information. This paper introduces a novel multimodal image fusion method mainly for improving contrast and as well as edge information. Primary step of this algorithm is to resize source images. The 3×3 sharpen filter and morphology hat transform are applied separately on resized IR image and VI image. DWT transform has been used to produce "low-frequency" and "high-frequency" sub-bands. "Filters based mean-weighted fusion rule" and "Filters based max-weighted fusion rule" are newly introduced in this algorithm for combining "low-frequency" sub-bands and "high-frequency" sub-bands respectively. Fused image reconstruction is done with IDWT. Proposed method has outperformed and shown improved results in subjective manner and objectively than similar existing techniques.

Segmentation of Online Freehand Sketching Based on Speed Feature

We present a novel method to extract speed feature points for segmenting hand-drawn strokes into geometric primitives. The method consists of three steps. Firstly, the input strokes are classified into uniform and nonuniform speed strokes, representing a stroke drawn at relatively constant or uneven speeds, respectively. Then, a sharpening filter is used to enhance the peak features of the uniform speed strokes. Finally, a three-threshold technique that uses the average speed of the pen and its upper and lower deviations is used to extract speed feature points of strokes. We integrate the proposed method into our freehand sketch recognition (FSR) system to improve its robustness to support multiprimitive strokes. Through a user study with 8 participants, we demonstrate that the proposed method achieves higher segmentation efficiency in finding speed feature points than the existing method based on a single speed threshold.

Use of noise reduction filters on stereo images for improving the accuracy and quality of the digital elevation model

Recently, there has been increasing attention to the digital elevation model (DEM) because of its ability to learn the Earth’s surface’s topography. Freely accessible DEMs such as the Cartosat-1, shuttle radar topography mission (SRTM), light detection and ranging, and advanced spaceborne thermal emission and reflection radiometer DEM, contain large vertical errors. The errors are aggravated over multifaceted geography and cannot rectify microgeographic deviations in the moderately flat landscape. As high-accuracy DEMs have limited availability, dated low-accuracy DEMs are still used in various models, specifically in the data-sparse areas. However, it is necessary to enhance the quality of these DEMs before their use in the geomorphometric analysis. We aim to investigate the effect of noise reduction filters on DEM’s accuracy and quality. The noise reduction filters such as weighted average filter, median filter, sharpen filter, Lee sigma filter, and local sigma filter are used for noise reduction in the stereo images. The quality and accuracy of the generated DEMs are further improved by selecting an optimum number of tie points in the image matching process. The effects are assessed by correlating the surface profiles for the final DEM obtained and SRTM DEM as a reference with a resolution of 1 arc sec.

WITHDRAWN: Features of urban architecture design based on ZYNQ embedded digital image processing

The city is excellent modern buildings are an excellent historical and cultural heritage, and its conservation studies include two important things: standard-setting and protection and management strategies. With the embedded digital image processing is becoming an important part of our lives, and image processing has reached its level. These systems have created many applications, such as processing in the field of image filtering and image processing. System implementation is completed with the low-cost Zed Board ZYNQ Seven thousand Field Programmable Gate Arrays (FPGA) and Raspberry-Pi. After various image processing filters, the output will be the input image. Get a Gaussian filter, a sharpening filter, and an average filter and display these results on your Video Graphics Array (VGA) monitor. Edge detection and grayscale operations are performed on Lena's image. Chip usage is detected. Using grid-based models and software algorithms derived from image processing, effective methods have been developed to measure geometric parameters and predict radiation exchange. The proposed system is generating synthetic urban textures and integrating cellular automata. The results show that a grid-based city model can provide information for planning and design. This system accelerates the detection of lung and lung disease and can be used as a single procedure or as part of other biomedical imaging systems. Image processing systems rely on bilateral filtration, edge recognition, multi-limits, image separation, text image processing, and image labeling.

Edge detection for weed recognition in lawns

The rapid propagation of weeds is a major issue for turfgrass management (both ornamental and sports turf). While pesticides can ensure weed eradication, they pose a risk to human health and the environment. In this context, the early detection of weeds can allow a dramatic reduction in the amount of pesticide required. Here we present the use of edge detection techniques to identify the presence of these invasive plants in ornamental lawns and sports turf. Regarding the former, images from small experimental plots in the facilities of IMIDRA were used while images for the latter were taken on a golf course. Up to 12 different filters for edge detection were tested on the images collected. Aggregation techniques, with a range of cell values, were applied to the results of the three most effective filters (sharpening (I), sharpening (II), and Laplacian) to minimise the number of false positives. After the tests with different cell sizes, two filters were selected for more in-depth analysis. Box plots were selected to define the best cell size and identify the filter with the best performance. The sharpening (I) filter and the aggregation technique with the minimum value and a cell size of 10 offered the best results. Finally, we determined the most appropriate threshold value on the basis of the number of false positives, false negatives, and derived indexes (Precision, Recall, and F1-Score). A threshold of 78 gave the best performance. The results achieved with this methodology differed slightly between ornamental and sports turf.

The Diagnosis Of Diabetic Retinopathy By Means Of Transfer Learning And Fine-Tuned Dense Layer Pipeline

Diabetes is a global disease that occurs when the body is disabled pancreas to secrete insulin to convert the sugar to power in the blood. As a result, some tiny blood vessels on the part of the body, such as the eyes, are affected by high sugar and cause blocking blood flow in the vessels, which is called diabetic retinopathy. This disease may lead to permanent blindness due to the growth of new vessels in the back of the retina causing it to detach from the eyes. In 2016, 387 million people were diagnosed with Diabetic retinopathy, and the number is growing yearly, and the old detection approach becomes worse. Therefore, the purpose of this paper is to computerize the old method of detecting different classes of DR from 0-4 according to severity by given fundus images. The method is to construct a fine-tuned deep learning model based on transfer learning with dense layers. The used models here are InceptionV3, VGG16, and ResNet50 with a sharpening filter. Subsequently, InceptionV3 has achieved 94% as the highest accuracy among other models.

Image recoloring using specular decomposition and in-painting

ABSTRACT Image recoloring using wavelet transform, in-painting and dichromatic reflection model is proposed. Detailed wavelet bands are passed through a sharpening filter whereas the approximation band is passed through dichromatic reflection model to extract body and specular coefficients. In-painting and white illumination minimization using erosion and Gaussian blur further enhance the image quality. The proposed technique produces better results, both qualitatively and quantitatively, in terms of edge preservation and minimization of noise, artefacts in contrast to state-of-art existing techniques.