Visibility Enhancement of Scene Images Degraded by Foggy Weather Condition: An Application to Video Surveillance

:In recent years,video surveillance application played a significant role in our daily lives.Images taken during foggy and haze weather conditions for video surveillance application lose their authenticity and hence reduces the visibility.The reason behind visibility enhancement of foggy and haze images is to help numerous computer and machine vision applications such as satellite imagery,object detection,target killing,and surveillance.To remove fog and enhance visibility,a number of visibility enhancement algorithms and methods have been proposed in the past.However,these techniques suffer from several limitations that place strong obstacles to the real world outdoor computer vision applications.The existing techniques do not perform well when images contain heavy fog,large white region and strong atmospheric light.This research work proposed a new framework to defog and dehaze the image in order to enhance the visibility of foggy and haze images.The proposed framework is based on a Conditional generative adversarial network(CGAN)with two networks;generator and discriminator,each having distinct properties.The generator network generates fog-free images from foggy images and discriminator network distinguishes between the restored image and the original fog-free image.Experiments are conducted on FRIDA dataset and haze images.To assess the performance of the proposed method on fog dataset,we use PSNR and SSIM,and for Haze dataset use e,r−,andσas performance metrics.Experimental results shows that the proposed method achieved higher values of PSNR and SSIM which is 18.23,0.823 and lower values produced by the compared method which are 13.94,0.791 and so on.Experimental results demonstrated that the proposed framework Has removed fog and enhanced the visibility of foggy and hazy images.

Role of surveillance imaging and endoscopy in colorectal cancer follow-up:Quality over quantity?

Colorectal cancer(CRC) is a prevalent disease and represents a major cause of morbidity and mortality in the developed world. Intensive post-treatment surveillance is routinely recommended by major expert groups for early stage(Ⅱ and Ⅲ) CRC survivors because previous meta-analyses showed a modest, but significant survival benefit. This practice has been recently challenged based on data emerging from several large phase Ⅲ randomized trials that demonstrated a lack of survival benefit from intensive surveillance strategies. In addition,findings from cost-effectiveness analyses of such an approach are inconsistent.Data on real-world practice, specifically adherence to these follow-up guidelines,are also limited. The debate is especially controversial in resected stage Ⅳ patients where there are currently no clear guidelines for follow-up. In an era of personalized medicine, there may be a shift towards a more risk-adapted approach to better define the optimal follow-up strategy. In this article, we review the evidence and highlight the role of surveillance in CRC survivors.

Role of abdominal ultrasound for the surveillance follow-up of pancreatic cystic neoplasms: a cost-effective safe alternative to the routine use of magnetic resonance imaging

BACKGROUND Patients with pancreatic cystic neoplasms(PCN), without surgical indication at the time of diagnosis according to current guidelines, require lifetime imagebased surveillance follow-up. In these patients, the current European evidencedbased guidelines advise magnetic resonance imaging(MRI) scanning every 6 mo in the first year, then annually for the next five years, without reference to any role for trans-abdominal ultrasound(US). In this study, we report on our clinical experience of a follow-up strategy of image-based surveillance with US, and restricted use of MRI every two years and for urgent evaluation whenever suspicious changes are detected by US.AIM To report the results and cost-efficacy of a US-based surveillance follow-up for known PCNs, with restricted use of MRI.METHODS We retrospectively evaluated the records of all the patients treated in our institution with non-surgical PCN who received follow-up abdominal US and restricted MRI from the time of diagnosis, between January 2012 and January 2017. After US diagnosis and MRI confirmation, all patients underwent US surveillance every 6 mo for the first year, and then annually. A MRI scan was routinely performed every 2 years, or at any stage for all suspicious US findings.In this communication, we reported the clinical results of this alternative followup, and the results of a comparative cost-analysis between our surveillance protocol(abdominal US and restricted MRI) and the same patient cohort that has been followed-up in strict accordance with the European guidelines recommended for an exclusive MRI-based surveillance protocol.RESULTS In the 5-year period, 200 patients entered the prescribed US-restricted MRI surveillance follow-up. Mean follow-up period was 25.1 ± 18.2 mo. Surgery was required in two patients(1%) because of the appearance of suspicious features at imaging(with complete concordance between the US scan and the on-demand MRI). During the follow-up, US revealed changes in PCN appearance in 28 patients(14%). These comprised main pancreatic duct dilatation(n = 1), increased size of the main cyst(n = 14) and increased number of PNC(n = 13). In all of these patients, MRI confirmed US findings, without adding more information.The bi-annual MRI identified evolution of the lesions not identified by US in only11 patients with intraductal papillary mucinous neoplasms(5.5%), largely consisting of an increased number of very small PCN(P = 0.14). The overall mean cost of surveillance, based on a theoretical use of the European evidenced-based exclusive MRI surveillance in the same group of patients, would have been 1158.9± 798.6 € per patient, in contrast with a significantly lower cost of 366.4 ± 348.7 €(P < 0.0001) incurred by the US-restricted MRI surveillance used at our institution.CONCLUSION In patients with non-surgical PCN at the time of diagnosis, US surveillance could be a safe complementary approach to MRI, delaying and reducing the numbers of second level examinations and therefore reducing the costs.

Role of serial multiparametric magnetic resonance imaging in prostate cancer active surveillance

AIM:To examine whether addition of 3T multiparametric magnetic resonance imaging(mp MRI)to an active surveillance protocol could detect aggressive or progressive prostate cancer.METHODS:Twenty-three patients with low risk disease were enrolled on this active surveillance study,all of which had Gleason score 6 or less disease.All patients had clinical assessments,including digital rectal examination and prostate specific antigen(PSA)testing,every 6 mo with annual 3T mp MRI scans with gadolinium contrast and minimum sextant prostate biopsies.The MRI images were anonymized of patient identifiers and clinical information and each scan underwentradiological review without the other results known.Descriptive statistics for demographics and follow-up as well as the sensitivity and specificity of mp MRI to identify prostate cancer and progressive disease were calculated.RESULTS:During follow-up(median 24.8 mo)11 of 23 patients with low-risk prostate cancer had disease progression and were taken off study to receive definitive treatment.Disease progression was identified through upstaging of Gleason score on subsequent biopsies for all 11 patients with only 2 patients also having a PSA doubling time of less than 2 years.All 23 patients had biopsy confirmed prostate cancer but only 10 had a positive index of suspicion on mp MRI scans at baseline(43.5% sensitivity).Aggressive disease prediction from baseline mpM RI scans had satisfactory specificity(81.8%)but low sensitivity(58.3%).Twentytwo patients had serial mp MRI scans and evidence of disease progression was seen for 3 patients all of whom had upstaging of Gleason score on biopsy(30% specificity and 100% sensitivity).CONCLUSION:Addition of mp MRI imaging in active surveillance decision making may help in identifying aggressive disease amongst men with indolent prostate cancer earlier than traditional methods.

Surveillance and diagnosis of hepatocellular carcinoma: A systematic review

BACKGROUND Hepatocellular carcinoma (HCC) appears in most of cases in patients with advanced liver disease and is currently the primary cause of death in this population.Surveillance of HCC has been proposed and recommended in clinical guidelines to obtain earlier diagnosis,but it is still controversial and is not accepted worldwide.AIM To review the actual evidence to support the surveillance programs in patients with cirrhosis as well as the diagnosis procedure.METHODS Systematic review of recent literature of surveillance (tools,interval,cost-benefit,target population) and the role of imaging diagnosis (radiological non-invasive diagnosis,optimal modality and agents) of HCC.RESULTS The benefits of surveillance of HCC,mainly with ultrasonography,have been assessed in several prospective and retrospective analysis,although the percentage of patients diagnosed in surveillance programs is still low.Surveillance of HCC permits diagnosis in early stages allows better access to curative treatment and increases life expectancy in patients with cirrhosis.HCC is a tumor with special radiological characteristics in computed tomography and magnetic resonance imaging,which allows highly accurate diagnosis without routine biopsy confirmation.The actual recommendation is to perform biopsy only in indeterminate nodules.CONCLUSION The evidence supports the recommendation of performing surveillance of HCC in patients with cirrhosis susceptible of treatment,using ultrasonography every 6 mo.The diagnosis evaluation of HCC can be established based on noninvasive imaging criteria in patients with cirrhosis.

Evolving screening and surveillance techniques for Barrett's esophagus

Barrett's esophagus(BE) is a change in the esophageal lining and is known to be the major precursor lesion for most cases of esophageal adenocarcinoma(EAC).Despite an understanding of its association with BE for many years and the falling incidence rates of squamous cell carcinoma of the esophagus, the incidence for EAC continues to rise exponentially. In association with this rising incidence, if the delay in diagnosis of EAC occurs after the onset of symptoms,then the mortality at 5 years is greater than 80%. Appropriate diagnosis and surveillance strategies are therefore vital for BE. Multiple novel optical technologies and other advanced approaches are being utilized to assist in making screening and surveillance more cost effective. We review the current guidelines and evolving techniques that are currently being evaluated.

Clinical impact of surveillance for head and neck cancer in patients with esophageal squamous cell carcinoma

To evaluate the clinical impact of surveillance for head and neck (HN) region with narrow band imaging (NBI) in patients with esophageal squamous cell carcinoma (ESCC).METHODSSince 2006, we introduced the surveillance for HN region using NBI for all patients with ESCC before treatment, and each follow-up. The patients with newly diagnosed stage I to III ESCC were enrolled and classified into two groups as follows: Group A (no surveillance for HN region); between 1992 and 2000), and Group B (surveillance for HN region with NBI; between 2006 and 2008). We comparatively evaluated the detection rate of superficial head and neck squamous cell carcinoma (HNSCC), and the serious events due to metachronous advanced HNSCC during the follow-up.RESULTSA total 561 patients (group A: 254, group B: 307) were enrolled. Synchronous superficial HNSCC was detected in 1 patient (0.3%) in group A, and in 12 (3.9%) in group B (P = 0.008). During the follow up period, metachronous HNSCC were detected in 10 patients (3.9%) in group A and in 30 patients (9.8%) in group B (P = 0.008). All metachronous lesions in group B were early stage, and 26 patients underwent local resection, however, 6 of 10 patients (60%) in group A lost their laryngeal function and died with metachronous HNSCC.CONCLUSIONSurveillance for the HN region by using NBI endoscopy increase the detection rate of early HNSCC in patients with ESCC, and led to decrease serious events related to advanced metachronous HNSCC.

Surveillance of patients following surgery with curative intent for colorectal cancer

Surveillance after resection of colorectal cancer with curative intent is an important component of post- operative care. Clinical review, imaging, colonoscopy, and cost to the community are among significant issues to consider in planning a surveillance regime. This review aims to identify the available evidence for the use of surveillance and its individual components. The literature pertaining to follow-up of patients following potentially curative surgery for colorectal cancer was reviewed in order to formulate a summary of the wide range of clinical practice. There is evidence of improved survival of patients undergoing more intense follow-up compared with those having minimal surveillance, with an estimated overall 5-year gain of up to 10%. The efficacy of individual components of follow-up regimes remains unclear, but an overall package of 'intensive' follow-up including clinical review, liver imaging, and colonoscopy appears to be of benefit. It is cost-effective and can be specialist or community-based.

Colorectal cancer surveillance in inflammatory bowel disease: A critical analysis

Colonoscopic surveillance is advocated in patients with inflammatory bowel disease(IBD) for detection of dys-plasia. There are many issues regarding surveillance in IBD: the risk of colorectal cancer seems to be de-creasing in the majority of recently published studies, necessitating revisions of surveillance strategy; surveil-lance guidelines are not based on concrete evidence; commencement and frequency of surveillance, cost-effectiveness and adherence to surveillance have been issues that are only partly answered. The traditional technique of random biopsy is neither evidence-based nor easy to practice. Therefore, highlighting abnormal areas with newer technology and biopsy from these areas are the way forward. Of the newer technology, digital mucosal enhancement, such as high-definition white light endoscopy and chromoendoscopy(with magnification) have been incorporated in guidelines. Dyeless chromoendoscopy(narrow band imaging) has not yet shown potential, whereas some forms of digital chromoendoscopy(i-Scan more than Fujinon intelligent color enhancement) have shown promise for colonoscopic surveillance in IBD. Other techniquessuch as autofluorescence imaging, endomicroscopy and endocytoscopy need further evidence. Surveillance with genetic markers(tissue, serum or stool) is at an early stage. This article discusses changing epidemiology of colorectal cancer development in IBD and critically evaluates issues regarding colonoscopic surveillance in IBD.

Outcomes of combination MRI-targeted and transperineal template biopsy in restaging low-risk prostate cancer for active surveillance

Objective:Active surveillance(AS)offers a strategy to reduce overtreatment and now is a widely accepted treatment option for low-risk prostate cancer.An ideal tool for risk-stratification would detect aggressive cancers and exclude such men from taking up AS in the first place.We evaluate if a combination of transperineal template biopsy with magnetic resonance imaging(MRI)-targeted biopsy identifies significant prostate cancer amongst men initially diagnosed with low-risk prostate cancer.Methods:This prospective,single-blinded study included men with low-risk prostate cancer(D’Amico’s Criteria)diagnosed on conventional transrectal ultrasound-guided biopsy.Patients first underwent multiparametric MRI of the prostate
6 weeks after initial biopsy.Each suspicious lesion is mapped and assigned a Prostate Imaging Reporting and Data System(PIRADS)score.Template biopsy is first performed with the surgeon blinded to MRI findings followed by MRI-targeted biopsy using a robotic transperineal biopsy platform.Results:The age of the 19 men included is 65.4±4.9 years(mean±SD).Prostate specific antigen(PSA)at diagnosis and at the time of transperineal biopsy were comparable(7.3±1.7 ng/mL and 7.0±1.8 ng/mL,p Z 0.67),so were prostate volumes(34.2±8.9 mL and 32.1±13.4 mL,p Z 0.28).MRI-targeted biopsy had a higher percentage of cancer detection per core compared to template biopsy(11.7%vs.6.5%,p Z 0.02),this was more than 3 times superior for Gleason 7 disease(5.9%vs.1.6%,p<0.01).Four of 18(22.2%)patients with MRI lesions had significant disease with MRI-targeted biopsy alone.Three of 19 patients(15.8%)had significant disease with template biopsy alone.In combination,both techniques upclassified five patients(26.3%),all of whom underwent radical prostatectomy.Whole mount histology confirmed tumour location and grade.All six patients with PIRADS 5 lesions had cancer detected(66.6%significant disease).Conclusion:A combination of MRI-targeted and template biopsy may optimally risk-classify“low-risk”patients diagnosed on initial conventional transrectal ultrasonography(TRUS)prostate biopsy.

OBJECT-BASED SUPER RESOLUTION FOR INTELLIGENT VISUAL SURVEILLANCE VIDEO

Construction of high resolution images from low resolution sequences is often im- portant in surveillance applications. In this letter, an affine based multi-scale block-matching image registration algorithm is first proposed. The images to be registered are divided into overlapped blocks of different size according to its motions. The Least Square (LS) image reg- istration algorithm is extended to match the blocks. Then an object based Super Resolution (SR) scheme is designed, the Maximum A Priori (MAP) super resolution algorithm is extended to enhance the resolution of the interest objects. Experimental results show that the proposed multi-scale registration method provides more accurate registration between frames. Further more, the object based super resolution scheme shows an enhanced performance compared with the traditional MAP method.

Smart Deep Learning Based Human Behaviour Classification for Video Surveillance

Real-time video surveillance system is commonly employed to aid security professionals in preventing crimes.The use of deep learning(DL)technologies has transformed real-time video surveillance into smart video surveillance systems that automate human behavior classification.The recognition of events in the surveillance videos is considered a hot research topic in the field of computer science and it is gaining significant attention.Human action recognition(HAR)is treated as a crucial issue in several applications areas and smart video surveillance to improve the security level.The advancements of the DL models help to accomplish improved recognition performance.In this view,this paper presents a smart deep-based human behavior classification(SDL-HBC)model for real-time video surveillance.The proposed SDL-HBC model majorly aims to employ an adaptive median filtering(AMF)based pre-processing to reduce the noise content.Also,the capsule network(CapsNet)model is utilized for the extraction of feature vectors and the hyperparameter tuning of the CapsNet model takes place utilizing the Adam optimizer.Finally,the differential evolution(DE)with stacked autoencoder(SAE)model is applied for the classification of human activities in the intelligent video surveillance system.The performance validation of the SDL-HBC technique takes place using two benchmark datasets such as the KTH dataset.The experimental outcomes reported the enhanced recognition performance of the SDL-HBC technique over the recent state of art approaches with maximum accuracy of 0.9922.

Detection of Objects in Motion—A Survey of Video Surveillance

Video surveillance system is the most important issue in homeland security field. It is used as a security system because of its ability to track and to detect a particular person. To overcome the lack of the conventional video surveillance system that is based on human perception, we introduce a novel cognitive video surveillance system (CVS) that is based on mobile agents. CVS offers important attributes such as suspect objects detection and smart camera cooperation for people tracking. According to many studies, an agent-based approach is appropriate for distributed systems, since mobile agents can transfer copies of themselves to other servers in the system.

基于深度学习的监控视频图像增强处理方法

传统的图像增强处理方法常存在图像清晰度较低、失真度较高的问题。因此,提出基于深度学习的增强处理方法。在获取监控视频图像信息后,对图像进行序列对齐处理。通过图像量化处理的方式显示单个图像中的关键像素点和邻近像素点之间的分布。采用深度学习技术,用8个方向的梯度算子提取反射图像的方向梯度,得到细节增强的反射图像,再通过Retinex反变换完成增强处理。设计对比实验,证明该方法可以改善监视视频的图像增强效果,获得更清楚的景物图片。

煤矿井下散射环境雾化图像的气室仿真方法

在粉尘环境中,井下监控图像会出现雾化退化现象,而在井下难以进行环境控制实验开展图像雾化的研究。因此,文章利用相对平均梯度定义图像雾化程度、分析了井下粉尘粒径构成、提出了井下粉尘配比方法、设计并搭建仿真气室模拟井下雾化图像。结果表明:利用平均梯度定义的雾化程度优于其他常用评价参数,其评价结果与人眼判断结果相符,而井下图像的雾化退化程度可以达到80%,此时图像的细节信息基本被雾化影响所掩盖;井下粉尘样本的测试结果显示:99%的井下粉尘粒径小于30μm,粒径小于10μm的粉尘的粒子数占比约为93%;利用仿真气室生成的雾化图像不仅可以涵盖井下图像雾化程度范围,相较于雾化算法生成的结果,具有更准确地模拟粉尘下的光散射扩散效果的优点。

Monitoring of hepatocellular carcinoma

Screening for hepatocellular carcinoma in patients at risk is an evidence-based approach;however,adherence to the monitoring protocol recommended by international guidelines is difficult.Hence,there is a need to use the best screening options and refine the selection of patients at risk in the future.

基于改进YOLOv7的码头作业人员检测算法

广角监控图像中人员目标检测对于码头智能安防具有重要意义。针对传统YOLOv7算法在码头广角监控图像识别中,存在小目标特征提取能力弱、人员检测准确率低等问题,研究了基于改进YOLOv7的码头作业人员检测算法。为提升人员目标多尺度特征的检测性能及鲁棒性,设计了平衡码头人员分类与定位任务的上下文解耦(task-specific context decoupling,TSCODE)结构并联合聚集-分发机制(gather-and-distribute,GD),增强网络多尺度特征融合能力;为增强网络对作业人员等小目标的特征提取能力,在主干网络末端引入了基于双层路由注意力机制(bi-level routing attention,BRA)的视觉transformer模型(BRA-ViT),捕捉小目标人员的位置、方向与跨通道等信息;为提升检测速度并保持检测精度,提出了基于slim-neck的颈部层网络轻量化方法,降低参数量与计算量;为降低漏检率与误检率,引入了基于最小点距离的交并比损失函数(minimum-point-distance-based intersection over union,MPDIoU)计算边界框的坐标预测损失,提升边界框回归的准确性与计算效率。为验证算法效果,采集白天、夜晚不同时段下码头前沿、堆场、卡口等场景的广角监控图像,构造标注数据集并设计消融与对比实验。实验结果显示:所提算法对码头作业人员检测的平均准确率为90.6%,平均检测速度为39 fps;与Faster R-CNN、SSD、YOLOv3、YOLOv5、YOLOv7、YOLOv8等算法相比,其平均准确率分别提升了13.8%、15.8%、8.5%、5.2%、2.7%和3.5%,平均检测速度与基准YOLOv7算法性能相当。所提算法对码头作业人员识别具有较高的检测精度与检测速度,满足码头安防场景中作业人员检测准确性与实时性的要求。

面向视频侦查应用的退化人像GFP深度复原技术

在视频侦查工作中时常遇到视频人像质量过低难以辨识的情况,而GFP等人像深度复原方法只适用于单张人像的复原。为此,本文提出一种基于GFP的监控视频人像复原技术,将GFP方法扩展至视频侦查应用,便于及时锁定犯罪嫌疑人,提高破案效率。首先对监控视频进行视频分帧、人像裁剪对齐、倾斜透视校正等预处理操作,然后使用GFP方法对预处理后得到的人像进行深度复原,最后经过逆处理将复原人像整合成高质量人像视频。在大量经过预处理后的模拟退化人像和无参考退化人像测试集上进行对比实验,结果表明,GFP方法在主观视觉效果和FID、PSFR、SSIM、NIQE等客观量化指标上均优于其他人像深度复原方法,相较于其他人像深度复原方法,能够更有效地复原复杂应用场景下的低质量退化人像,更适用于视频侦查应用;通过使用YTF视频人像数据集进行对比测试实验,结果显示,本文所提出的添加预处理与逆处理过程的基于GFP的监控视频人像复原技术,对于低质量视频人像有更加优秀的复原效果。

多尺度域视频监控图像光照分量智能校正方法

针对视频监控图像中存在的清晰度不足和光照不均匀等问题,提出了多尺度域视频监控图像光照分量智能校正方法。应用背景差分法与差值图二值化法对视频监控图像进行预处理,以减少噪声干扰。通过融合TH变换和BH变换来加强图像多尺度细节特征;利用控制因子控制锐化强度,以凸显细节特征,避免出现图像伪影情况。对图像光照分量进行自适应调节,利用二维伽马函数计算画面亮度平均值,以防止出现过度校正的情况,从而实现智能校正。测试结果表明,采用该方法校正的图像更清晰,画面光照分量均匀,图像失真较少,视觉效果较好。

用于视频监控系统的生物特征识别技术标准化研究

为了提升生物特征识别技术在视频监控系统中的应用效果,对比阐述了生物特征识别系统与视频监控系统中存在语义差异的常用术语,并介绍了具备生物特征识别功能的视频监控系统的处理流程和重要组件,分析了影响生物特征识别技术在视频监控系统中使用的主要因素,提出如何从主要硬件及软件部署、参考数据库规范以及性能评价等方面进行标准化考量,最后从视频标注的规范性以及典型生物特征识别标准在视频监控系统中的进一步应用推广两个层面进行了展望,从而提升该领域的标准化整体水平。