Gaussian Smoothing Filter Research Articles

Sub-minute acquisition with deep learning-based image filter in the diagnosis of colorectal cancers using total-body 18F-FDG PET/CT

BackgroundThis study aimed to retrospectively evaluate the feasibility of total-body 18F-FDG PET/CT ultrafast acquisition combined with a deep learning (DL) image filter in the diagnosis of colorectal cancers (CRCs).MethodsThe clinical and preoperative imaging data of patients with CRCs were collected. All patients underwent a 300-s list-mode total-body 18F-FDG PET/CT scan. The dataset was divided into groups with acquisition durations of 10, 20, 30, 60, and 120 s. PET images were reconstructed using ordered subset expectation maximisation, and post-processing filters, including a Gaussian smoothing filter with 3 mm full width at half maximum (3 mm FWHM) and a DL image filter. The effects of the Gaussian and DL image filters on image quality, detection rate, and uptake value of primary and liver metastases of CRCs at different acquisition durations were compared using a 5-point Likert scale and semi-quantitative analysis, with the 300-s image with a Gaussian filter as the standard.ResultsAll 34 recruited patients with CRCs had single colorectal lesions, and the diagnosis was verified pathologically. Of the total patients, 11 had liver metastases, and 113 liver metastases were detected. The 10-s dataset could not be evaluated due to high noise, regardless of whether it was filtered by Gaussian or DL image filters. The signal-to-noise ratio (SNR) of the liver and mediastinal blood pool in the images acquired for 10, 20, 30, and 60 s with a Gaussian filter was lower than that of the 300-s images (P < 0.01). The DL filter significantly improved the SNR and visual image quality score compared to the Gaussian filter (P < 0.01). There was no statistical difference in the SNR of the liver and mediastinal blood pool, SUVmax and TBR of CRCs and liver metastases, and the number of detectable liver metastases between the 20- and 30-s DL image filter and 300-s images with the Gaussian filter (P > 0.05).ConclusionsThe DL filter can significantly improve the image quality of total-body 18F-FDG PET/CT ultrafast acquisition. Deep learning-based image filtering methods can significantly reduce the noise of ultrafast acquisition, making them suitable for clinical diagnosis possible.

Effect of Gaussian Smoothing Filter Size for CT-Based Attenuation Correction on Quantitative Assessment of Bone SPECT/CT: A Phantom Study.

This study aims to determine the effect of Gaussian filter size for CT-based attenuation correction (CTAC) on the quantitative assessment of bone SPECT. An experiment was performed using a cylindrical phantom containing six rods, of which one was filled with water and five were filled with various concentrations of K2HPO4 solution (120-960mg/cm3) to simulate different bone densities. 99mTc-solution of 207kBq/ml was also included within the rods. SPECT data were acquired at 120 views for 30s/view. CT for attenuation correction were obtained at 120 kVp and 100mA. Sixteen different CTAC maps processed with different Gaussian filter sizes (ranging from 0 to 30mm in 2mm increments) were generated. SPECT images were reconstructed for each of the 16 CTAC maps. Attenuation coefficients and radioactivity concentrations in the rods were compared with those in the water-filled rod without K2HPO4 solution as a reference. Gaussian filter sizes below 14-16mm resulted in an overestimation of radioactivity concentrations for rods with high concentrations of K2HPO4 (≥ 666mg/cm3). The overestimation of radioactivity concentration measurement was 3.8% and 5.5% for 666mg/cm3 and 960mg/cm3 K2HPO4 solutions, respectively. The difference in radioactivity concentration between the water rod and the K2HPO4 rods was minimal at 18-22mm. The use of Gaussian filter sizes smaller than 14-16mm caused an overestimation of radioactivity concentration in regions of high CT values. Setting the Gaussian filter size to 18-22mm enables radioactivity concentration to be measured with the least influence on bone density.

Liver fat analysis using optimized support vector machine with support vector regression.

Fatty liver disease is a common condition caused by excess fat in the liver. It consists of two types: Alcoholic Fatty Liver Disease, also called alcoholic steatohepatitis, and Non-Alcoholic Fatty Liver Disease (NAFLD). As per epidemiological studies, fatty liver encompasses 9% to 32% of the general population in India and affects overweight people. An Optimized Support Vector Machine with Support Vector Regression model is proposed to evaluate the volume of liver fat by image analysis (LFA-OSVM-SVR). The input computed tomography (CT) liver images are collected from the Chennai liver foundation and Liver Segmentation (LiTS) datasets. Here, input datasets are pre-processed using Gaussian smoothing filter and bypass filter to reduce noise and improve image intensity. The proposed U-Net method is used to perform the liver segmentation. The Optimized Support Vector Machine is used to classify the liver images as fatty liver image and normal images. The support vector regression (SVR) is utilized for analyzing the fat in percentage. The LFA-OSVM-SVR model effectively analyzed the liver fat from CT scan images. The proposed approach is activated in python and its efficiency is analyzed under certain performance metrics. The proposed LFA-OSVM-SVR method attains 33.4%, 28.3%, 25.7% improved accuracy with 55%, 47.7%, 32.6% lower error rate for fatty image classification and 30%, 21%, 19.5% improved accuracy with 57.9%, 46.5%, 31.76% lower error rate for normal image classificationthan compared to existing methods such as Convolutional Neural Network (CNN) with Fractional Differential Enhancement (FDE) (CNN-FDE), Fully Convolutional Networks (FCN) and Non-negative Matrix Factorization (NMF) (FCN-NMF), and Deep Learning with Fully Convolutional Networks (FCN) (DL-FCN).

An area-delay efficient single-precision floating-point multiplier for VLSI systems

The floating-point multiplier (FPM) is the most commonly used component in various image and signal processing applications. An area-delay efficient FPM design could be helpful for the development of high-performance VLSI computing systems. Therefore, this paper presents the analysis of single-precision FPM to find the possibilities for the reduction in delay and area. From the analysis, it is found that the delay of FPM depends on the delay of mantissa multiplication and exponent normalization units. Based on the above observations, in this paper, the delay and area efficient mantissa multiplication and exponent addition-cum-normalization units are proposed. Further, using these components, the single-precision FPM structure is proposed. Synthesis results show that the proposed FPM design offers a saving of 15.3% in delay and 10.8% in area compared to the best available existing FPM design. Besides, the proposed FPM consumes 13.0% less power-delay-product and involves 24.4% less area-delay-product in comparison to the best available existing design. Finally, effectiveness of the proposed single precision floating-point multiplier is presented in Gaussian smoothing filter of the image processing application.

State of Health Estimation of Lithium-Ion Batteries in Electric Vehicles Based on Regional Capacity and LGBM

Battery state of health (SOH) estimation is a prerequisite for battery health management and is vital for second-life utilization. Existing techniques implemented in well-controlled experimental conditions fail to reflect complex working conditions during actual vehicular operation. In this article, a novel SOH estimation method for battery systems in real-world electric vehicles (EVs) is presented by combing results of regional capacity calculation and a light gradient boosting machine (LGBM) model. The LGBM model is used to capture the relationship between battery degeneration and influential factors based on datasets from real-world EVs. The regional capacity, which is calculated through incremental capacity analysis with a Gaussian smoothing filter, is utilized to reflect the battery degradation level while ensuring high flexibility and applicability. Accumulated mileage, average charging current, average charging temperature, and start and end of SOC values are chosen as influential factors for model establishment. The effectiveness, complexity, superiority, and robustness of the proposed method are verified using data from real-world EVs. Results indicate accurate SOH estimation can be achieved with an average absolute error of only 0.89 Ah, where the MAPE and RMSE of the test vehicles are 2.049% and 1.153%, respectively.

Computer-Vision-Based Statue Detection with Gaussian Smoothing Filter and EfficientDet

Smart tourism is a developing industry, and numerous nations are planning to establish smart cities in which technology is employed to make life easier and link nearly everything. Many researchers have created object detectors; however, there is a demand for lightweight versions that can fit into smartphones and other edge devices. The goal of this research is to demonstrate the notion of employing a mobile application that can detect statues efficiently on mobile applications, and also improve the performance of the models by employing the Gaussian Smoothing Filter (GSF). In this study, three object detection models, EfficientDet—D0, EfficientDet—D2 and EfficientDet—D4, were trained on original and smoothened images; moreover, their performance was compared to find a model efficient detection score that is easy to run on a mobile phone. EfficientDet—D4, trained on smoothened images, achieves a Mean Average Precision (mAP) of 0.811, an mAP-50 of 1 and an mAP-75 of 0.90.

Evolution algorithm of parametric active contour model based on Gaussian smoothing filter

Evolution algorithm of parametric active contour model based on Gaussian smoothing filter

A machine learning classification methodology for Raman Hyperspectral imagery based on auto‐encoders

AbstractRaman spectroscopy is a very sensitive, non‐destructive optical technique that is able to characterize the composition of a material. Although it is a very useful technique, its signals are rather complex and therefore require a lot of knowledge and skills to understand their meaning. This complexity increases further, when it is combined with spatial mapping, creating a three‐dimensional dataset named Raman hyperspectral imagery. The common methodology of analysing such a dataset is to use a generic supervised classification algorithm (i.e., non‐negative least square) combined with reference data. If reference data are not available, unsupervised algorithms are used to perform spectral clusterization of similar areas. This is typically achieved using standard PCA and k‐means methods. The most common problem of processing that significantly affects the results of analysis is that a raw Raman Signal consists of three sources: Raman scattering, photoluminescence and noise. The contribution weights of those components are not standard and are changing based on the sample's chemistry, which makes it very difficult to be solved. This creates the necessity for pre‐processing methods that can split this raw signal to its components. The goals of this project are to improve the unmixing performance using AI and to achieve a successful split of the raw signal into its three components. Through this study, various unsupervised methods are explored to identify the method that fits best to our microscopic RHSI data sets. Our focus is on using autoencoders for the data analysis and on a Gaussian smoothing filter combined with polynomial fittings to split the data. Results show that the selection of a classification methodology depends greatly on the application, but when a large RHSI dataset is available, the newly proposed splitting technique combined with an autoencoder classification presents the best solution.

Effect of Gaussian filtered images on Mask RCNN in detection and segmentation of potholes in smart cities.

Accidents have contributed a lot to the loss of lives of motorists and serious damage to vehicles around the globe. Potholes are the major cause of these accidents. It is very important to build a model that will help in recognizing these potholes on vehicles. Several object detection models based on deep learning and computer vision were developed to detect these potholes. It is very important to develop a lightweight model with high accuracy and detection speed. In this study, we employed a Mask RCNN model with ResNet-50 and MobileNetv1 as the backbone to improve detection, and also compared the performance of the proposed Mask RCNN based on original training images and the images that were filtered using a Gaussian smoothing filter. It was observed that the ResNet trained on Gaussian filtered images outperformed all the employed models.

High-accuracy mode recognition method in orbital angular momentum optical communication system

Vortex optical communication has been a hot research field in recent years. A key step is mode recognition in the orbital angular momentum (OAM) free-space optical (FSO) communication system. In this article, we propose an OAM mode recognition method based on image recognition technology, which uses the interferogram between the vortex beam and the Gaussian beam to identify the OAM mode. In order to resist the influence of atmospheric turbulence on the recognition accuracy, we added a Gaussian smoothing filter into the recognition process. Moreover, we used random phase screens to generate interferogram sets at distances of 1 km and 2 km. The verification result shows that the proposed scheme produces high identification accuracy for the distorted optical field. The average accuracy can reach 100% and 87.78% under the conditions of medium- and strong-turbulence levels, respectively. It is anticipated that these results might be helpful for improving the reliability of the OAM-FSO communication system in the future.

A recognition method of multispectral images of soybean canopies based on neural network

Multispectral images of soybean canopies can reflect plant physiological information and growth status effectively, which is of great significance for soybean high-quality breeding, scientific cultivation, and fine management. At present, it is uneven of the gray level difference of the soybean multispectral images occurred in the leaf edge, and is also small of the gray level difference between the target and the background, resulting in inaccurate recognition of the soybean canopies from the multispectral images. Thus, a multispectral images' recognition method of soybean canopies was proposed based on the neural network. First, the method of Gaussian smoothing filter was used to preprocess the raw soybean multispectral images (green light, near-infrared, red light, red edge, and visible light), which maintained the leaf edge details of the soybean multispectral image. Second, the feedforward neural network model was established to extract the canopy region in the soybean multispectral images. In addition, image morphology operation was used to improve the recognition effects of the soybean canopy. Finally, four quantitative indexes were used to evaluate the experimental results. The results showed that the average effective segmentation rate of the proposed method was 91.69%, the under-segmentation rate was reduced by 33.34%, and the over-segmentation rate was reduced by 48.43%. The difference between the pixel average entropy of the proposed method and the standard canopy image was only 0.2295. The research results can provide not only reliable data for further analysis of physiological and ecological indexes of the soybean canopy, but also technical support for multispectral image recognition of other crop canopies.

Determination of 3D pore network structure of freeze-dried maltodextrin

This article describes and discusses the determination of the pore network structure from X-ray microcomputed tomography images of freeze-dried maltodextrin based on image processing. Maltodextrin is used because of its relevance for research and pharmaceutical applications. Image processing is based on conventional methods for image filtering and binarization, which are specifically the Gaussian filter, anisotropic diffusion filter, and Otsu thresholding as well as adaptive thresholding. It is shown, that the combination of the different image processing methods has only a minor impact on the final results of the estimated porosity. More critical is the segmentation of the overall void space into individual pores using the watershed algorithm, especially if the markers of the Euclidean distance map are not accurately defined. Oversegmentation of the void space into too many small pores was obviated by the use of a Gaussian smoothing filter on the Euclidean distance map. The approximation of the structure by spherical or cylindrical geometrical objects was addressed in further discussions as an essential step for the determination of the mathematical pore network model. It was found that the choice of the specific geometrical object has a major impact on the final pore size distribution of the generated pore networks. For the case of maltodextrin, with long and narrow pores, the approximation by segmented cylinders, replacing the individual pore subvolumes, yields the best agreement. The computed pore network structures are further investigated in regard of the correlation of vapor transport regimes and structure variation. It is shown that the small differences obtained from the various combinations of image processing and network reconstruction methods have a negligible impact if temperature and pressure are low (T = −18 °C, P = 10 Pa) and constant.

Incorporation of anatomical MRI knowledge for enhanced mapping of brain metabolism using functional PET

Functional positron emission tomography (fPET) imaging using continuous infusion of [18F]-fluorodeoxyglucose (FDG) is a novel neuroimaging technique to track dynamic glucose utilization in the brain. In comparison to conventional static or dynamic bolus PET, fPET maintains a sustained supply of glucose in the blood plasma which improves sensitivity to measure dynamic glucose changes in the brain, and enables mapping of dynamic brain activity in task-based and resting-state fPET studies. However, there is a trade-off between temporal resolution and spatial noise due to the low concentration of FDG and the limited sensitivity of multi-ring PET scanners. Images from fPET studies suffer from partial volume errors and residual scatter noise that may cause the cerebral metabolic functional maps to be biased. Gaussian smoothing filters used to denoise the fPET images are suboptimal, as they introduce additional partial volume errors. In this work, a post-processing framework based on a magnetic resonance (MR) Bowsher-like prior was used to improve the spatial and temporal signal to noise characteristics of the fPET images. The performance of the MR guided method was compared with conventional denosing methods using both simulated and in vivo task fPET datasets. The results demonstrate that the MR-guided fPET framework denoises the fPET images and improves the partial volume correction, consequently enhancing the sensitivity to identify brain activation, and improving the anatomical accuracy for mapping changes of brain metabolism in response to a visual stimulation task. The framework extends the use of functional PET to investigate the dynamics of brain metabolic responses for faster presentation of brain activation tasks, and for applications in low dose PET imaging.

A saliency-based multiscale approach for infrared and visible image fusion

The ideal fusion of the infrared image and visual image should integrate complete bright features of the infrared image, and preserve original visual information of the visual image as much as possible. To this end, we propose a multi-scale decomposition fusion method based on saliency. In particular, the saliency detection and a Gaussian smoothing filter are first employed to decompose source images into salient layers, detail layers and base layers. Then we adopt a nonlinear function to calculate the weight coefficient to fuse salient layers and highlight the target. Subsequently, we use a fusion rule based on phase congruency for fusion of detail layers so that the details could be retained better than the traditional “max-absolute” fusion rule. Experiments show that the proposed method can achieve better fusion effect than the state-of-the-art methods qualitatively and quantitatively. Moreover, for the ill-illumination fused image, in order to get better visual effect, we further propose a contrast enhancement algorithm based on total variation minimization. Experiments show that the proposed method can enhance the contrast and retain details of the source images well.

Energy-Efficient Gaussian Filter Design Using Novel Low-Complexity Accuracy Reconfigurable Reverse Carry Adder

This paper first presents novel approximate reverse carry full adder (ARFA) circuits where the weight of the input carry is more than that of the output carry. Using the proposed ARFAs and conventional full adders, new approximate reverse carry propagate adder (ARCAs) architectures are presented where the carry propagates in a counter manner. The proposed ARCA provides trade-off in accuracy and performance depending on the number of ARFAs used. Further, novel accuracy reconfigurable reverse carry propagate adders are proposed that provide approximate sum with high energy efficiency and can also deliver 100% accurate sum with small performance penalty. For comparative analysis, the proposed and existing adders are modeled in Verilog and synthesized via Synopsys Design Compiler with 65 nm PDK. The synthesis results of the proposed 16-bit ARCA-II show 9.93% reduced energy consumption over the existing. Finally, the Gaussian smoothing filter (GSF) embedded with the proposed adder shows 12.15% reduced energy consumption with an acceptable image quality (1.64 dB loss in PSNR) over the GSFs with the existing adder.

An Efficient Implementation of Low-Latency Two-Dimensional Gaussian Smoothing Filter using Approximate Carry-Save Adder

The two-dimensional Gaussian smoothing filter (2D-GSF) is one of the most useful techniques in image processing. Since the 2D-GSF requires high computational resources, its efficient design and implementation are critical in real-time processing purposes. Approximate computing is a new method that can be used to increase the performance of 2D Gaussian filter design with low computing overhead on field-programmable gate arrays (FPGAs). This study aims to provide a low-latency Gaussian filter architecture on FPGA such that it can be used in real-time processing applications. In this regard, accurate and approximate carry-save adders (CSAs) have been used in adder tree-based Gaussian filters. In our proposed method, we use two approximation steps: in the first step, we use an approximation structure named Speed–Power–Area–Accuracy for Gaussian filter design and in the second stage, we use approximate CSAs to convert adder-tree structures that are used in Gaussian filter, and as a result, we have significantly reduced the delay. The results of simulation and implementation show that the latency has reduced in a 3[Formula: see text] 3 2D-GSF architecture up to 22% using proposed accurate CSAs and 45% using proposed approximate CSAs, compared to existing Gaussian filters with an adder tree structure.

Reconfigurable Carry Look-Ahead Adder Trading Accuracy for Energy Efficiency

The modern portable devices exhibiting multimedia applications demand higher energy efficient signal processing due to limited battery size. The approximate adders have shown a remarkable energy-efficiency over the accurate adders for error tolerant applications. In this paper, three novel approximate carry look-ahead adder (ACLA) architectures are proposed. These approximate ACLAs are achieved by simplifying the Boolean expression of carry generation logic such that the probability of error is small while eliminating number of logic gates. Further, a novel accuracy reconfigurable CLA (Re-CLA) that provides desired quality/accuracy in the given energy budget is proposed. The post layout synthesis results using Synopsys IC Compiler of the proposed adders are computed and analysed against the existing adders. These results demonstrate 37.67%, 18.21%, 18.14% and 15.92% reduction in energy consumption by the proposed 8-bit ACLA-I, -II, -III and -IV adders respectively over the existing approximate adder. Further, the proposed 8-bit and 16-bit Re-CLAs require only 1.92% and 7.08% more energy over the existing CLA for achieving accuracy reconfigurability. Finally, the synthesis results of the Gaussian smoothing filters embedded with the proposed adders show higher energy efficiency with acceptable image quality over the state-of-the-art adder architectures.

LoBA: A Leading One Bit Based Imprecise Multiplier for Efficient Image Processing

Several applications such as signal processing, multimedia and big data analysis exhibit computational error tolerance. This tolerance can be exploited to achieve efficient designs by sacrificing accuracy. Therefore, approximate computing presents a new design paradigm that smashes the traditional belief of error-free computations and provides efficient design with quality metrics specific to an application. The multiplication operation significantly determines the performance of the core due to the compute intensive operation. Therefore, this paper proposes a novel leading one bit based approximate (LoBA) multiplier architecture that selects k-bits from n-bit inputs (k ≤ n/4) based on leading one bit (LOB) and then computes approximate product based of these small input. The accuracy is further improved by selecting next k-bits based on LOB position and considering the partial product for computing final product. Four imprecise LoBA multipliers are presented that provide trade-off between accuracy and performance. Finally, the effectiveness of the proposed architectures is shown over the existing multipliers as standalone arithmetic unit and in the application by implementing Gaussian smoothing filters. The proposed 16-bit LoBA0 and LoBA1 designs reduce power consumption by 64.2% and 32.9%, respectively over the existing multiplier architecture.

Connected component analysis integrated edge based technique for automatic vehicular license plate recognition framework

Automatic vehicle license plate recognition (AVLPR) aims at extracting the region that contains the information of vehicle license number out of an image data and then identifying the characters apart from the human intervention. This study proposed an effective AVLPR framework where detection, segmentation and recognition of various shaped license plates have been focused. For both proper visual perception and computational processing, a pre-processing technique including grey-scaling conversion combined with close arithmetic-based dilation has been defined. Both vertical and horizontal edge densities have been enumerated by kernel matrices which enable robustness in detecting various shaped and sized license plates. For better detection of candidate region, the vertical and horizontal energy mapping features combined with Gaussian smoothing filter have been used to enable detection of license plates from both high definition and lower resolution images under various illumination conditions and crowded background. For ensuring a better character segmentation rate which is the prerequisite for higher recognition rate, a blob assessment method has been defined integrated with connected component analysis. With 400 vehicle images having varying pixels, the proposed algorithm achieves 96.5, 95.6 and 94.4% accuracy, respectively, in identifying, segmenting and recognising the plate number.

Application Of A Template Matching Algorithm In The Detection Of The Railroad Track

As the most important part of urban public transportation system, railway and subway are undertaking extremely important tasks. Manual inspection is sometimes difficult to detect track damage in time. Based on the impulse excitation of the rail, the undamaged rail and the damaged rail are distinguished according to the characteristics of its vibration signal. In the process of feature extraction, the pulse signal is intercepted, and the threshold is taken after gaussian smoothing filter. After multiple processing, the feature template is formed. Through the template matching algorithm proposed in this paper, the rail excitation pulse signal can be identified and the rail damage state can be determined. And the simulation display and control platform based on real object is made. After many tests, the method is proved to be effective.