Region-Aware Fashion Contrastive Learning for Unified Attribute Recognition and Composed Retrieval

Clothing attribute recognition has become an essential technology,which enables users to automatically identify the characteristics of clothes and search for clothing images with similar attributes.However,existing methods cannot recognize newly added attributes and may fail to capture region-level visual features.To address the aforementioned issues,a region-aware fashion contrastive language-image pre-training(RaF-CLIP)model was proposed.This model aligned cropped and segmented images with category and multiple fine-grained attribute texts,achieving the matching of fashion region and corresponding texts through contrastive learning.Clothing retrieval found suitable clothing based on the user-specified clothing categories and attributes,and to further improve the accuracy of retrieval,an attribute-guided composed network(AGCN)as an additional component on RaF-CLIP was introduced,specifically designed for composed image retrieval.This task aimed to modify the reference image based on textual expressions to retrieve the expected target.By adopting a transformer-based bidirectional attention and gating mechanism,it realized the fusion and selection of image features and attribute text features.Experimental results show that the proposed model achieves a mean precision of 0.6633 for attribute recognition tasks and a recall@10(recall@k is defined as the percentage of correct samples appearing in the top k retrieval results)of 39.18 for composed image retrieval task,satisfying user needs for freely searching for clothing through images and texts.

Detection of broken manhole cover using improved Hough and image contrast

The damage or loss of urban road manhole covers may cause great risk to residents＇ lives and property if they cannot be discovered in time. Most existing research recommendations for solving this problem are difficult to implement. This paper proposes an algorithm that combines the improved Hough transform and image comparison to identify the damage or loss of the manhole covers in complicated surface conditions by using existing urban road video images. Focusing on the pre-processed images, the edge contour tracking algorithm is applied to find all of the edges. Then with the improved Hough transformation, color recognition and image matching algorithm, the manhole cover area is found and the change rates of the manhole cover area are calculated. Based on the threshold of the change rates, it can be determined whether there is potential damage or loss in the manhole cover. Compared with the traditional Hough transform, the proposed method can effectively improve the processing speed and reduce invalid sampling and accumulation. Experimental results indicate that the proposed algorithm has the functions of effective positioning and early warning in the conditions of complex background, different perspectives, and different videoing time and conditions, such as when the target is partially covered.

Novel registration algorithm for 3-D images captured from multiple views of object surface

A novel algorithm of 3-D surface image registration is proposed. It makes use of the array information of 3-D points and takes vector/vertex-like features as the basis of the matching. That array information of 3-D points can be easily obtained when capturing original 3-D images. The iterative least-mean-squared （LMS） algorithm is applied to optimizing adaptively the transformation matrix parameters. These can effectively improve the registration performance and hurry up the matching process. Experimental results show that it can reach a good subjective impression on aligned 3-D images. Although the algorithm focuses primarily on the human head model, it can also be used for other objects with small modifications.

DEVELOPMENT OF MOVING TARGET DETECTION AND IMAGING BY AIRBORNE SAR

The detection and ima ging of moving targets based on airborne synthetic aperture radar (SAR) is a cru cial technique for the modern radar. Firstly, the mathematical model of SAR ech o signal which comes from moving targets is constructed. Based on this model, th e features of moving target imaging are introduced and the effects of target mov ement to SAR imaging are analyzed. Then the development and the status of this t echnique are reviewed in detail. Finally, some frontiers of this field are point ed out.

An improved detail enhancement algorithm based on difference curvature and contrast field

The gradient image is always sensitive to noise in image detail enhancement. To overcome this shortage, an improved detail enhancement algorithm based on difference curvature and contrast field is proposed. Firstly, the difference curvature is utilized to determine the amplification coefficient instead of the gradient. This new amplification function of the difference curvature takes more neighboring points into account, it is therefore not sensitive to noise. Secondly, the contrast field is nonlinearly amplified according to the new amplification coefficient. And then, with the enhanced contrast field, we construct the energy functional. Finally, the enhanced image is reconstructed by the variational method. Experimental results of standard testing image and industrial X-ray image show that the proposed algorithm can perform well on increasing contrast and sharpening edges of images while suppressing noise at the same time.

Joint inversion of P-wave velocity and Vp–Vs ratio: imaging the deep structure in NE Japan

A new inversion scheme is presented to obtain three-dimensional images of P-wave velocity（Vp） and P–S-wave velocity ratio（Vp/Vs） using P- and S-phase pairs, i.e., the same source–receiver pairs for the P- and S-wave arrival-time data. The S-wave velocity（Vs） was separately inverted using the S-phase arrival times. The earthquake hypocenters were simultaneously relocated in the joint inversion. The method considers the Vp/Vs anomaly as a model parameter in the inversion. The proposed method thus provides a more robust calculation of the Vp/Vs anomaly than the conventional method of dividing Vp by Vs. The method also takes into account the ray path difference between P- and S-waves, and hence yields a less biased Vp–Vs ratio than the method of inverting S–P-wave data for Vp and Vp/Vs anomalies under the assumption of identical P and S ray paths. The proposed method was used to image the crust and upper mantle in northeastern（NE） Japan taking advantage of a large number of high-quality arrival times of P- and S-wave source–receiver pairs. The inverted structures suggest that the subducting slab of the Pacific plate is an inclined zone of high-Vp and Vs anomalies with low Vp/Vs perturbation. The mantle wedge is characterized by low-Vp, low-Vs, and high-Vp/Vs anomalies at shallow depths beneath active volcanoes. These features are also observed at greater depths in the back-arc region. Although these features have been previously reported, the Vp/Vs anomaly pattern obtained in this study shows much less scatter and is much better correlated with the seismic velocity perturbation patterns than previous studies. The proposed method can be used, in conjunction with velocity anomaly patterns, to quantify thermal processes associated with plate subduction.

Review of Remotely Sensed Imagery Classification Patterns Based on Object-oriented Image Analysis

With the wide use of high-resolution remotely sensed imagery, the object-oriented remotely sensed informa- tion classification pattern has been intensively studied. Starting with the definition of object-oriented remotely sensed information classification pattern and a literature review of related research progress, this paper sums up 4 developing phases of object-oriented classification pattern during the past 20 years. Then, we discuss the three aspects of method- ology in detail, namely remotely sensed imagery segmentation, feature analysis and feature selection, and classification rule generation, through comparing them with remotely sensed information classification method based on per-pixel. At last, this paper presents several points that need to be paid attention to in the future studies on object-oriented RS in- formation classification pattern: 1) developing robust and highly effective image segmentation algorithm for multi-spectral RS imagery; 2) improving the feature-set including edge, spatial-adjacent and temporal characteristics; 3) discussing the classification rule generation classifier based on the decision tree; 4) presenting evaluation methods for classification result by object-oriented classification pattern.

Effects of RapidEye Imagery's Red-edge Band and Vegetation Indices on Land Cover Classification in an Arid Region

Land cover classification(LCC) in arid regions is of great significance to the assessment, prediction, and management of land desertification. Some studies have shown that the red-edge band of RapidE ye images was effective for vegetation identification and could improve LCC accuracy. However, there has been no investigation of the effects of RapidE ye images' red-edge band and vegetation indices on LCC in arid regions where there are spectrally similar land covers mixed with very high or low vegetation coverage information and bare land. This study focused on a typical inland arid desert region located in Dunhuang Basin of northwestern China. First, five feature sets including or excluding the red-edge band and vegetation indices were constructed. Then, a land cover classification system involving plant communities was developed. Finally, random forest algorithm-based models with different feature sets were utilized for LCC. The conclusions drawn were as follows: 1) the red-edge band showed slight contribution to LCC accuracy; 2) vegetation indices had a significant positive effect on LCC; 3) simultaneous addition of the red-edge band and vegetation indices achieved a significant overall accuracy improvement(3.46% from 86.67%). In general, vegetation indices had larger effect than the red-edge band, and simultaneous addition of them significantly increased the accuracy of LCC in arid regions.

Flotation bubble seed image filling algorithm based on boundary point features

Bubble seed image filling is an important prerequisite for the image segmentation of flotation bubble that can be used to improve flotation automatic control. These common image filling algorithms in dealing with complex bubble image exists under-filling and over-filling problems. A new filling algorithm based on boundary point feature and scan lines ~PFSL） is proposed in the paper. The filling a｜gorithm describes these boundary points of image objects by means of chain codes. The features of each boundary point, including convex points, concave points, left points and right points, are defined by the point＇s entrancing chain code and leaving chain code. The algorithm firstly finds out all double-matched boundary points based on the features of boundary points, and fill image objects by these double-matched boundary points on scan lines. Experimental results of bubble seed image filling show that under-filling and over-filling problem can be eliminated by the proposed algorithm.

Multi-phase post-mortem CT-angiography： a pathologic correlation study on cardiovascular sudden death

Multi-phase post-mortem CT-angiography （MPMCTA） has the great potential to increase the quality of the postmortem investigation, especially in the area of sudden death; however, its role as routine complement to the pathology toolbox is still questioned as it needs to be further standardized. The aim of this study is to investigate the contribution of MPMCTA in cases of sudden unexplained death in adults and in particular in sudden cardiovascular death. Sixty-eight sudden unexpected deaths of adults were investigated at our institution between 2012 and 2013. Ten cases underwent MPMCTA and autopsy and were included in the study. Before the angiographic step by complete filling of the vascular system, prior to any manipulation of the body, a non-contrast CT-scan was carried out. Image reconstructions were performed on a CT workstation （Vitrea） and two radiologists experienced with post mortem imaging interpreted the MPMCTA findings. In all 10 cases, we could state a good correlation between combination of post-mortem CT and MPMCTA and autopsy procedures, confirming a high diagnostic sensitivity. With this case series we want to illustrate the advantages offered by performing MPMCTA when facing a sudden death, regardless of specific suspicion for acute coronary syndrome or other vascular or ischemic disease.

Life-threatening Spontaneous Retroperitoneal Haemorrhage:Role of Multidetector CT-angiography for the Emergency Management

Objective To evaluate the role of multidetector computed tomography-angiography(MDCTA) for management of life-threatening spontaneous retroperitoneal haemorrhage(SRH).Methods A retrospective analysis of the medical records of all SRH patients admitted to our hospital from January 1,2005 to November 31,2014 was performed.SRH was defined as unrelated to invasive procedures,surgery,trauma,abdominal tumor or aortic aneurysm.Life-threatening SRH was defined as hemodynamic instability with hypotension.Sixteen cases met the inclusive criteria.Patients' characteristics,clinical presentation,imaging of MDCTA,treatment modalities,morbidity,and follow-up data were analyzed.Results The group of 16 patients included 11 men and 5 women,with a mean age of 51.5 ± 16.4years.All patients underwent MDCTA scan examinations,and active contrast media(CM) extravasation were found in 9 patients.Angiography was performed for 8 patients in the CM extravasation group.Bleeding vessels were found in 7 patients who underwent successful embolization.One patient in the CM extravasation group did not undergo angiography and died of bleeding.Seven patients without active CM extravasation underwent conservative management.Bleeding was controlled for all these 7 patients,and 1 patient died of multiple organ failure.Fourteen patients survived from SRH were successfully discharged.One patient died of heart failure during follow-up.Conclusion MDCTA was useful in the diagnosis of SRH and proved to be predictive for the outcome of conservative treatment and early intervention.

Design of Content-Based Retrieval System in Remote Sensing Image Database

To retrieve the object region efficaciously from massive remote sensing image database, a model for content-based retrieval of remote sensing image is given according to the characters of remote sensing image application firstly, and then the algorithm adopted for feature extraction and multidimensional indexing, and relevance feedback by this model are analyzed in detail. Finally, the contents intending to be researched about this model are proposed.

Mammography combined with breast dynamic contrast-enhanced-magnetic resonance imaging for the diagnosis of early breast cancer

Objective The aim of this study was to investigate the application of mammography combined with breast dynamic contrast-enhanced magnetic resonance imaging(DCE-MRI) for the diagnosis of early breast cancer. Methods Mammography and DCE-MRI were performed for 120 patients with breast cancer(malignant, 102; benign; 18). Results The sensitivity of mammography for early diagnosis of breast cancer was 66.67%, specificity was 77.78%, and accuracy was 68.33%. The sensitivity of MRI for early diagnosis of breast cancer was 94.12%, specificity was 88.89%, and accuracy was 93.33%. However, the sensitivity of mammography combined with DCE-MRI volume imaging with enhanced water signal(VIEWS) scanning for early diagnosis of breast cancer was 97.06%, specificity was 94.44%, and accuracy was 96.67%. Conclusion Mammography combined with DCE-MRI increased the sensitivity, specificity, and accuracy of diagnosing early breast cancer.

Diagnostic accuracy of real-time tissue elastography for breast cancer:a meta-analysis

Objective The present study aimed to determine the accuracy of real-time tissue elastography （RTE） for the diagnosis of breast cancer. Methods The search was conducted in the PubMed, Web of Science, Cochrane Library, and China Biology Medicine databases from inception through December 31, 2014, without language restrictions. The meta-analysis was conducted using STATA version 12.0 and Meta-Disc version 1.4. We calculated the summary statistics for sensitivity （Sen）, specificity （Spe）, positive and negative likelihood ratio （LR＋/LR–）, diagnostic odds ratio （DOR）, and summary receiver operating characteristic （SROC） curve. Results Ten studies that met al inclusion criteria were included in the meta-analysis. A total of 608 ma-lignant breast lesions and 1292 benign breast tumors were assessed. Al breast lesions were histological y confirmed after RTE. The pooled Sen was 0.83 （95% CI = 0.79–0.86）; the pooled Spe was 0.86 （95% CI = 0.84–0.88）. The pooled LR＋ was 9.87 （95% CI = 2.66–36.71）; the pooled LR– was 0.20 （95% CI = 0.17–0.23）. The pooled DOR of RTE for the diagnosis of breast cancer was 62.21 （95% CI = 33.88–114.24）. The area under the SROC curve was 0.9334 （standard error = 0.00125）. We found no evidence of publica-tion bias （t = –0.57, P = 0.582）. Conclusion RTE may have high diagnostic accuracy for the dif erential diagnosis of benign and malig-nant breast tumors. RTE may be a good tool for breast cancer diagnosis.

RESEARCH ON COLOR CONSTANCY UNDER OPEN ILLUMINATION CONDITIONS

Color information plays a key role in the research fields of object recognition and image retrieval. However, the actual color varies by the conditions of illumination, especially the open natural daylight. Four different color constancy schemes are proposed in the paper to minimize the effects of open illumination conditions. (1) The color constancy scheme based on the image statistics is proposed, which includes the color cast detection and removal. (2) The color constancy scheme based on the color temperature curve is proposed, which combines Gaussian model with linear fitting to estimate color temperature curve. (3) The color constancy scheme based on the double exposure theory is proposed, which is able to reproduce a color image under typical illumination. (4) According to the concepts of supervised learning, the supervised color constancy scheme is proposed. The transformation of color values from unknown illumination to typical illumination is solved by improved Support Vector Regression (SVR).

Recognition of Similar Weather Scenarios in Terminal Area Based on Contrastive Learning

In order to improve the recognition accuracy of similar weather scenarios(SWSs)in terminal area,a recognition model for SWS based on contrastive learning(SWS-CL)is proposed.Firstly,a data augmentation method is designed to improve the number and quality of weather scenarios samples according to the characteristics of convective weather images.Secondly,in the pre-trained recognition model of SWS-CL,a loss function is formulated to minimize the distance between the anchor and positive samples,and maximize the distance between the anchor and the negative samples in the latent space.Finally,the pre-trained SWS-CL model is fine-tuned with labeled samples to improve the recognition accuracy of SWS.The comparative experiments on the weather images of Guangzhou terminal area show that the proposed data augmentation method can effectively improve the quality of weather image dataset,and the proposed SWS-CL model can achieve satisfactory recognition accuracy.It is also verified that the fine-tuned SWS-CL model has obvious advantages in datasets with sparse labels.

A special growth manner of intrahepatic biliary cystadenoma

We report a case of a 56-year-old woman with intrahepatic biliary cystadenoma(IBC)accompanying a tumor embolus in the extrahepatic bile duct,who was admitted to our department on October 13,2008.Imaging showed an asymmetry dilation of the biliary tree,different bile signals in the biliary tree,a multiloculated lesion and an extrahepatic bile duct lesion with internal septation.A regular left hemihepatectomy en bloc was performed with resection of the entire tumor,during which a tumor embolus protruding into the extrahepatic bile duct and originating from biliary duct of segment 4 was revealed.Microscopically,the multiloculated tumor was confirmed to be a biliary cystadenoma with an epithelial lining composed of biliary-type cuboidal cells and surrounded by an ovarian-like stroma.An aggressive en bloc resection was recommended for the multiloculated lesion.Imaging workup,clinicians and surgeons need to be aware of this different presentation.

Diagnostic value of gadobenate dimeglumine-enhanced hepatocyte-phase magnetic resonance imaging in evaluating hepatic fibrosis and hepatitis

AIM To evaluate the diagnostic value of gadobenate dimeglumine(Gd-BOPTA)-enhanced hepatocyte-phase magnetic resonance imaging(MRI) in evaluating hepatic fibrosis and hepatitis.METHODS Hepatocyte-phase images of Gd-BOPTA-enhanced MRI were retrospectively evaluated in 76 patients with chronic liver disease. These patients were classified into five groups according to either the histopathological fibrosis stage(S0-S4) or the histopathological hepatitis grade(G0-G4). The relative enhancement ratio(RE) of the liver parenchyma in the T1-vibe sequence was calculated by measuring the signal intensity before(SI pre) and 90 min after(SI post) intravenous injection of Gd-BOPTA using the following formula: RE =(SI post-SI pre)/SI pre. One-way analysis of variance was used to compare the difference between the relative RE in the hepatocyte phase(REh) and the stage of hepatic fibrosis and the grade of hepatitis. Pearson's productmoment correlation analysis was used to evaluate the relationship between the REh and the levels of serologic liver functional parameters.RESULTS According to histopathological hepatic fibrosis stage, the 76 patients were classified into five groups: 16 in S0, 15 in S1, 21 in S2, 9 in S3, and 15 in S4 group. According to histopathological hepatitis grade, the 76 patients were also classified into five groups: 0 in G0, 44 in G1, 22 in G2, 8 in G3, and 2 in G3 group. With regard to the stage of hepatic fibrosis, REh showed significant differences between the S2 and S3 groups and between the S2 and S4 groups(P < 0.05), but no significant difference was observed between the other groups. With regard to the grade of hepatitis, REh showed significant differences between the G1 and G2 groups and between the G1 and G4 groups(P < 0.05), but no significant difference was observed between the other groups. Increased REh showed correlations with decreased serum levels of TB, ALT and AST(P < 0.05). CONCLUSION To some extent, measuring the REh using Gd-BOPTAenhanced MRI might be a noninvasive technique for assessing the stage of hepatic fibrosis. This method is able to differentiate no/mild hepatitis from advanced hepatitis. TB, ALT and AST levels can predict the degree of liver enhancement in the hepatocyte phase of Gd-BOPTA-enhanced MRI.

Getting pre-stack time migration travel times from the single square root operator

Improving the focusing capability of pre-stack time migration allows the imaged section to reflect structural characteristics, depth, and interface shape and it is a key step for the preparation of the initial depth migration velocity model. The traditional symmetrical travel time equation is derived based on the assumption of a layered model. It is difficult to achieve the desired effect of focusing in media with strong lateral variation. The nonsymmetrical travel time equation based on Lie algebra and a pseudo-differential operator contains a lateral velocity derivative which can improve the focusing capability even in strongly lateral variable media and also the computation precision of the weight coefficients for relative amplitude preservation. Compared with the symmetrical methods, the nonsymmetrical method is more effective. In this paper, we describe several key steps of nonsymmetric pre-stack travel time calculation and present some test results using synthetic and real data.

Primary clear cell carcinoma in the liver: CT and MRI findings

AIM: To retrospectively analyze the computed tomography (CT) and magnetic resonance imaging (MRI) appearances of primary clear cell carcinoma of the liver (PCCCL) and compare the imaging appearances of PCCCL and common type hepatocellular carcinoma (CHCC) to determine whether any differences exist between the two groups. METHODS: Twenty cases with pathologically proven PCCCL and 127 cases with CHCC in the Second Affiliated Hospital of Sun Yat-sen University were included in this study. CT or MRI images from these patients were retrospectively analyzed. The following imaging findings were reviewed: the presence of liver cirrhosis, tumor size, the enhancement pattern on dynamic contrast scanning, the presence of pseudo capsules, tumor rupture, portal vein thrombosis and lymph node metastasis. RESULTS: Both PCCCL and CHCC were prone to occur in patients with liver cirrhosis, the association rate of liver cirrhosis was 80.0% and 78.7%, respectively (P > 0.05). The mean sizes of PCCCL and CHCC tumors were (7.28 ± 4.25) cm and (6.96 ± 3.98) cm, respectively. Small HCCs were found in 25.0% (5/20) of PCCCL and 19.7% (25/127) of CHCC cases. No significant differences in mean size and ratio of small HCCs were found between the two groups (P = 0.658 and 0.803, respectively). Compared with CHCC patients, PCCCL patients were more prone to form pseudo capsules (49.6% vs 75.0%, P = 0.034). Tumor rupture, typical HCC enhancement patterns and portal vein tumor thrombosis were detected in 15.0% (3/20), 72.2% (13/18) and 20.0% (4/20) of patients with PCCCL and 3.1% (4/127), 83.6% (97/116) and 17.3% (22/127) of patients with CHCC, respectively. There were no significant differences between the two groups (all P > 0.05). No patients with PCCCL and 2.4% (3/127) of patients with CHCC showed signs of lymph node metastasis (P > 0.05). CONCLUSION: The imaging characteristics of PCCCL are similar to those of CHCC and could be useful for differentiating these from other liver tumors (such as hemangioma and hepatic metastases). PCCCLs are more prone than CHCCs to form pseudo capsules.