Types Of Image Data Research Articles

DR|GRADUATE: Uncertainty-aware deep learning-based diabetic retinopathy grading in eye fundus images

Diabetic retinopathy (DR) grading is crucial in determining the adequate treatment and follow up of patient, but the screening process can be tiresome and prone to errors. Deep learning approaches have shown promising performance as computer-aided diagnosis (CAD) systems, but their black-box behaviour hinders clinical application. We propose DR|GRADUATE, a novel deep learning-based DR grading CAD system that supports its decision by providing a medically interpretable explanation and an estimation of how uncertain that prediction is, allowing the ophthalmologist to measure how much that decision should be trusted. We designed DR|GRADUATE taking into account the ordinal nature of the DR grading problem. A novel Gaussian-sampling approach built upon a Multiple Instance Learning framework allow DR|GRADUATE to infer an image grade associated with an explanation map and a prediction uncertainty while being trained only with image-wise labels. DR|GRADUATE was trained on the Kaggle DR detection training set and evaluated across multiple datasets. In DR grading, a quadratic-weighted Cohen's kappa (κ) between 0.71 and 0.84 was achieved in five different datasets. We show that high κ values occur for images with low prediction uncertainty, thus indicating that this uncertainty is a valid measure of the predictions' quality. Further, bad quality images are generally associated with higher uncertainties, showing that images not suitable for diagnosis indeed lead to less trustworthy predictions. Additionally, tests on unfamiliar medical image data types suggest that DR|GRADUATE allows outlier detection. The attention maps generally highlight regions of interest for diagnosis. These results show the great potential of DR|GRADUATE as a second-opinion system in DR severity grading.

The Impact of Emotional Music on Active ROI in Patients with Depression Based on Deep Learning: A Task-State fMRI Study

With the continuous development of science, more and more research results have proved that machine learning is capable of diagnosing and studying the major depressive disorder (MDD) in the brain. We propose a deep learning network with multibranch and local residual feedback, for four different types of functional magnetic resonance imaging (fMRI) data produced by depressed patients and control people under the condition of listening to positive- and negative-emotions music. We use the large convolution kernel of the same size as the correlation matrix to match the features and obtain the results of feature matching of 264 regions of interest (ROIs). Firstly, four-dimensional fMRI data are used to generate the two-dimensional correlation matrix of one person’s brain based on ROIs and then processed by the threshold value which is selected according to the characteristics of complex network and small-world network. After that, the deep learning model in this paper is compared with support vector machine (SVM), logistic regression (LR), k-nearest neighbor (kNN), a common deep neural network (DNN), and a deep convolutional neural network (CNN) for classification. Finally, we further calculate the matched ROIs from the intermediate results of our deep learning model which can help related fields further explore the pathogeny of depression patients.

A real-time magnetic resonance imaging study of cross-speaker variability in the production of /ɹ/

There is a longstanding interest in understanding the production mechanisms of the American English (AE) rhotic, /ɹ/ especially given both the wide intraspeaker variability and interspeaker heterogeneity in its articulation, and the resultant acoustics. A diversity of tongue shapes, from those creating a more anterior tongue-tip (“retroflex”) constriction to those with a more posterior bunched tongue dorsum constriction, has been observed with notable intraspeaker variation as a function of context; more specifically, a tendency for retroflexion preceding low and/or back vowels. The present study aims to analyze AE /ɹ/ production using real-time MRI videos (83 frames/s) where the target sound was produced in several contexts by 32 American English speakers (16 F/16 M). Initial findings in symmetric VCV production reveal that some of these speakers exclusively use a “retroflex” strategy, some a “bunched” strategy, and others both strategies, consistent with the previously observed tendencies. This observed variability, and the fact that multiple types of static and dynamic imaging data are available from the same speakers, makes this dataset ideal for further probing /ɹ/ production strategies in conjunction with speaker-specific characteristics of vocal tract structure, such as palate shape, and function, such as inter-articulator coordination. [Work supported by NIH and NSF.]

Development and Publication of Volcanic Base Map Data

Abstract. Japan has 111 active volcanoes accounting for about 7 percent of active volcanoes in the world and many volcanic disasters have been recorded so far. Geospatial Information Authority of Japan (GSI) proceeds to develop Volcanic Base Map, which depicts specific landforms of volcanoes precisely from 1979 to 2009 for the purpose of providing basic information for formulating various types of volcanic disaster prevention plans and volcano research. Although Volcanic Base Maps have been created by aerial photogrammetry for 31 active volcanoes throughout Japan as a paper map with a scale of 1:2,500 to 1:10,000, they have not been able to meet the needs of recent volcanic disaster prevention measures and have not been fully utilized in disaster emergency measures due to the following reasons: (1) There are undeveloped volcanoes because creation pace was limited for one volcano per year in the conventional creation method; (2) Considerable number of years has passed after its maintenance, so the latest terrain and infrastructure have not been updated; (3) It is impossible to cover the whole area of the volcano and the range of damage estimation because the development around the crater is mainly focused; (4) Basic information such as disaster prevention related facility information, which is necessary for the formulating volcanic disaster prevention plan and disaster emergency measures, has not been covered; (5) Since they were provided by a paper map for a fee, they were not suitable for advanced application using GIS.In order to solve the above issues and to promptly develop and provide maps of volcanoes, we have revised methods of making and acquisition criteria and have newly created Volcanic Base Map Data from 2017. Contour line data was created using digital stereoscopic machine in the conventional method of Volcanic Base Map, but it was changed to generate automatically by Digital Elevation Model of airborne laser survey. Also, we have changed to create features such as roads, buildings and disaster prevention-related facilities by using existing geospatial data maintained by GSI and materials collected from relevant organs. Therefore, the process of numerical mapping and associated photocontrol points surveying, pricking and aerial triangulation becomes omissible, and the number of creating maps can be increased. Although some of the designated evacuation shelters and heliports are not acquired in the conventional Volcanic Base Map, they are added as an acquisition item in Volcanic Base Map Data in order to make it easier to utilize to disaster prevention plan and emergency measures.Volcanic Base Map Data consists of three types of raster image data (base map, colour relief map and photomap, Figure 1), two types of vector data (Shapefile and Digital Mapping data) and GSI Tile data, is started to provide and publish on March 31, 2017 on the website of GSI and 14 volcanoes are provided as of October 2018. While Volcanic Base Map Data is possible to update in a very short time compared to conventional paper maps by maintaining and creating as numerical data, if there is a change in the contents of display such as disaster prevention related facilities and important infrastructures, it is necessary to consider a mechanism that can quickly update and publish maps. Furthermore, it is necessary to promote the products in each of the Volcano Disaster Management Councils and to continuously grasp the needs of users to be utilized more in the future.

The Practice of Deep Learning Methods in Biodiversity Information Collection

Deep learning is one machine learning method based on the layers used in artificial neural networks. The breakthrough of deep learning in classification led to its rapid application in speech recognition, natural language understanding, and image processing and recognition. In the field of biodiversity informatics, deep learning efforts are being applied in rapid species identification and counts of individuals identified based on image, audio, video, and other data types. However, deep learning methods hold great potential for application in all aspects of biodiversity informatics. We present a case study illustrating how to increase data collection quality and efficiency using well-established technology such as optical character recognition (OCR) and some image classification. Our goal is to image data from the scanned documents of various butterfly atlases, add species, specimens, collections, photographs and other relevant information, and build a database of butterfly images. Information collection involves image annotation and text-based descritpion input. Although the work of image annotation is simple, this process can be accelerated by deep learning-based target segmentation to make the selection process easier, such as changing box select to a double click. The process of information collection is complicated, involving input of species names, specimen collection, specimen description, and other information. Generally, there are many images in atlases, the text layout is rather messy, and overall OCR success is poor. Therefore, the measures we take are as follows: Step A: select the screenshot of the text and then call the OCR interface to generate the text material; Step B: proceed with NLP- (natural language processing) related processing; Step C: perform manual operations on the results, and introduce the NLP function again to this process; Step D: submit the result. The deep learning applications we integrated in our client tool include: target segmentation of the annotated image for automatic positioning and background removal, etc. to improve the quality of the image used for identification; making a preliminary judgment on various attributes of the labeled image and using the results to assist the automatic filling of relevant information in step B, including species information, specimen attributes (specimen image, nature photo, hand drawing pictures, etc.), insect stage (egg, adult, etc.); OCR in step A. target segmentation of the annotated image for automatic positioning and background removal, etc. to improve the quality of the image used for identification; making a preliminary judgment on various attributes of the labeled image and using the results to assist the automatic filling of relevant information in step B, including species information, specimen attributes (specimen image, nature photo, hand drawing pictures, etc.), insect stage (egg, adult, etc.); OCR in step A. Some simple machine learning methods such as k-nearest neighbor can be used to automatically determine gender, pose, and so on. While complex information such as collection place, time, and collector can be analyzed by deep learning-based NLP methods in the future. In our infomation collection process, ten fields are required to submit one single record. Of those, about 4-5 input fields can be dealt with the AI-assistant. It can thus be seen from the above process that deep learning has reduced the workload of manual information annotation by at least 30%. With improvements in accuracy, the era of using automatic information extraction robots to replace manual information annotation and collection is just around the corner.

The software and hardware architecture of the real-time protection of in-vessel components in JET-ILW

For the first time, the JET operation in deuterium–tritium (D–T) plasma, which is scheduled to take place on JET in 2020, will be performed in the ITER mix of plasma-facing component materials. In view of the preparation of the DT campaign (DTE2), several aspects of the plasma operation require significant improvements, such as a real-time protection of the first wall. The risk of damaging the metallic PFCs caused by beryllium melting or cracking of tungsten owing to thermal fatigue required a new reliable D–T compatible active protection system. Therefore, the future development of the JET real time first wall protection is focused on the D–T campaign and the ITER relevant conditions which may cause failure of camera electronics within the Torus hall. In addition to the technological aspect, the intensive preparation of the diverse software tools and real time algorithms for hot spot detection as well as alarm handling strategy required for the wall protection is in progress.This contribution describes the improved design, implementation, and operation of the near infrared (NIR) imaging diagnostic system of the JET-ILW plasma experiment and its integration into the existing JET protection architecture. To provide the reliable wall protection during the DTE2, two more sensitive logarithmic NIR camera systems equipped with new optical relays to take images and cameras outside of the biological shield have been installed on JET-ILW and calibrated with an in-vessel calibration light source (ICLS). Additionally, post-pulse data visualization and advanced analysis of all types of imaging data is provided by the new software framework JUVIL (JET users video imaging library). The formation of hot spots is recognized as a significant threat due to rapid surface temperature rise. Because it could trigger the protection system to stop a pulse, it is important to identify the mechanisms and conditions responsible for the formation of such hot spots. To address this issue the new software tool ‘Hotspot Editor’ has been developed.

Numerical simulations in 3-dimensions of reaction–diffusion models for brain tumour growth

ABSTRACTWe work with a well-known model of reaction–diffusion type for brain tumour growth and accomplish full 3-dimensional (3d) simulations of the tumour in time on two types of imaging data, the 3d Shepp–Logan head phantom image and an MRI T1-weighted brain scan from the Internet Brain Segmentation Repository. The source term is such that we have logistic growth. These simulations are obtained using standard finite difference approximations with novel calculations to increase speed and accuracy. Moreover, biological background to the model, its well-posedness together with a variational formulation are given. The variational formulation enable the feasibility of different derivations and modifications of the model.

Simulating permeabilities based on 3D image data of a layered nano-porous membrane

The macroscopic properties of materials are strongly influenced by their microstructure. Essential micro-structural information can be obtained from high-resolution 3D images, usually obtained by computed tomography. Here, a layered nano-porous ZrO2 Al2O3 membrane for fluid filtration is investigated, whose two finer layers can not be resolved properly by computed tomography. 3D image data of these structures are therefore obtained by the combination of focused ion beam sectioning and scanning electron microscopy. Reconstruction of porous structures from this type of image data and subsequent numerical simulation of flow properties within the reconstructed structure bear several sources of error. Using stochastic geometry models for the solid component, we examine more closely the error due to the typical anisotropic voxel setting as well as the error due to the also typical uneven sample shape.

PCA-based supervised identification of biological soil crusts in multispectral images

It was the aim of the method development to classify types of various biological soil crusts (biocrusts) using principle component analysis (PCA) on multispectral images. To address this aim, visible (RGB) and NIR images of bare sandy soil, algal and moss biocrusts were registered, per channel reflection values were determined using a calibration color chart on a pixel basis, and a PCA was performed on the unfolded RGB-NIR reflectance hypercubes (i.e. three-dimensional hypercubes were transformed into x.y × λ 2D-matrices with λ channels serving as variables for PCA). The classification approach was based on the hypothesis that biocrust types map specifically in PCA ordination plots, meaning that distinct regions in ordination plots may be assigned specifically to individual biocrust types. Reallocation of the pixels assigned to biocrust types to their respective image coordinates would then yield biocrust classification plots.•Allows manual selection of features or identification of given features in PCA ordination plots.•Fully permits the selection of relevant and omission of irrelevant, as well as identification of unknown classes.•It is not restricted to RGB-NIR multispectral data only, but may be applied to any type of multimodal imaging data.

Empirical cross sensor comparison of Sentinel-2A and 2B MSI, Landsat-8 OLI, and Landsat-7 ETM+ top of atmosphere spectral characteristics over the conterminous United States

Remote sensing landscape monitoring approaches frequently benefit from a dense time series of observations. To enhance these time series, data from multiple satellite systems need to be integrated. Landsat image data is a valuable 30-meter resolution source of spatial information to assess forest conditions over time. Together both operational Landsat satellites—7 and 8—provide a revisit frequency of 8 days at the equator. This moderate temporal frequency provides essential information to detect annual large area abrupt land cover changes. However, the ability to measure subtle and short lived intraseasonal changes is challenged by gaps in Landsat imagery at key points in time. The first Sentinel-2 satellite mission was launched by the European Space Agency in 2015. This moderate resolution data stream provides an opportunity to supplement the Landsat data record. The objective of this study is to assess the potential for integrating top of atmosphere Landsat and Sentinel 2 image data archived in the Google Earth Engine compute environment. In this paper we assess absolute and proportional differences in near-contemporaneous observations for six bands with comparable spectral response functions and spatial resolution between the Sentinel-2 Multi Spectral Instrument and Landsat Operational Land Imager and Enhanced Thematic Mapper Plus imagery. We assessed differences using absolute difference metrics and major axis linear regression between over 10,000 image pairs across the conterminous United States and present cross sensor transformation models. Major axis linear regression results indicate that Sentinel MSI data are as spectrally comparable to the two types of Landsat image data as the Landsat sensors are with each other. Root-mean-square deviation (RMSD) values ranging from 0.0121 to 0.0398 were obtained between MSI and Landsat spectral values, and RMSD values ranging from 0.0124 and 0.0372 were obtained between OLI and ETM+. Despite differences in their spatial, spectral, and temporal characteristics, integration of these datasets appears to be feasible through the application of bandwise linear regression corrections.

Automatic Assessment of Full Left Ventricular Coverage in Cardiac Cine Magnetic Resonance Imaging with Fisher Discriminative 3D CNN.

Cardiac magnetic resonance (CMR) images play a growing role in the diagnostic imaging of cardiovascular diseases. Full coverage of the left ventricle (LV), from base to apex, is a basic criterion for CMR image quality and necessary for accurate measurement of cardiac volume and functional assessment. Incomplete coverage of the LV is identified through visual inspection, which is time-consuming and usually done retrospectively in the assessment of large imaging cohorts. This paper proposes a novel automatic method for determining LV coverage from CMR images by using Fisher-discriminative three-dimensional (FD3D) convolutional neural networks (CNNs). In contrast to our previous method employing 2D CNNs, this approach utilizes spatial contextual information in CMR volumes, extracts more representative high-level features and enhances the discriminative capacity of the baseline 2D CNN learning framework, thus achieving superior detection accuracy. A two-stage framework is proposed to identify missing basal and apical slices in measurements of CMR volume. First, the FD3D CNN extracts high-level features from the CMR stacks. These image representations are then used to detect the missing basal and apical slices. Compared to the traditional 3D CNN strategy, the proposed FD3D CNN minimizes within-class scatter and maximizes between-class scatter. We performed extensive experiments to validate the proposed method on more than 5,000 independent volumetric CMR scans from the UK Biobank study, achieving low error rates for missing basal/apical slice detection (4.9%/4.6%). The proposed method can also be adopted for assessing LV coverage for other types of CMR image data.

Combinations of Multiple Neuroimaging Markers using Logistic Regression for Auxiliary Diagnosis of Alzheimer Disease and Mild Cognitive Impairment

Background: Multiple neuroimaging modalities have been developed providing various aspects of information on the human brain. Objective: Used together and properly, these complementary multimodal neuroimaging data integrate multisource information which can facilitate a diagnosis and improve the diagnostic accuracy. Methods: In this study, 3 types of brain imaging data (sMRI, FDG-PET, and florbetapir-PET) were fused in the hope to improve diagnostic accuracy, and multivariate methods (logistic regression) were applied to these trimodal neuroimaging indices. Then, the receiver-operating characteristic (ROC) method was used to analyze the outcomes of the logistic classifier, with either each index, multiples from each modality, or all indices from all 3 modalities, to investigate their differential abilities to identify the disease. Results: With increasing numbers of indices within each modality and across modalities, the accuracy of identifying Alzheimer disease (AD) increases to varying degrees. For example, the area under the ROC curve is above 0.98 when all the indices from the 3 imaging data types are combined. Conclusion: Using a combination of different indices, the results confirmed the initial hypothesis that different biomarkers were potentially complementary, and thus the conjoint analysis of multiple information from multiple sources would improve the capability to identify diseases such as AD and mild cognitive impairment.

Classification of Spatiotemporal Neural Activity Patterns in Brain Imaging Data

Various patterns of neural activity are observed in dynamic cortical imaging data. Such patterns may reflect how neurons communicate using the underlying circuitry to perform appropriate functions; thus it is crucial to investigate the spatiotemporal characteristics of the observed neural activity patterns. In general, however, neural activities are highly nonlinear and complex, so it is a demanding job to analyze them quantitatively or to classify the patterns of observed activities in various types of imaging data. Here, we present our implementation of a novel method that successfully addresses the above issues for precise comparison and classification of neural activity patterns. Based on two-dimensional representations of the geometric structure and temporal evolution of activity patterns, our method successfully classified a number of computer-generated sample patterns created from combinations of various spatial and temporal patterns. In addition, we validated our method with voltage-sensitive dye imaging data of Alzheimer’s disease (AD) model mice. Our analysis algorithm successfully distinguished the activity data of AD mice from that of wild type with significantly higher performance than previously suggested methods. Our result provides a pragmatic solution for precise analysis of spatiotemporal patterns of neural imaging data.

An improved data-hiding approach using skin-tone detection for video steganography

Securing the embedded data and diminishing the distortions in videos remain as challenging goals in data-hiding techniques. In this paper, a steganographic approach is proposed to minimize the probability of detection of the embedded image data in cover video objects. Human skin regions are considered as Regions of Interest (ROI) for embedding the secret data because the hidden data is not noticeable to human visual system (HVS). The proposed technique works with both 2-D and 3-D secret images of any size, which increases the adaptability of the proposed model to the various types of image data. Selective continuous pixel stacks are used to embed the secret data in the cover video object, which is performed over the third Discrete Wavelet Transform (DWT) component. The approximation coefficient has been entitled to be utilized for the purpose of embedding after the application of third Level DWT over the input video frame to enhance robustness and video quality. The frame matrix further undergoes the skin-map extraction to make the color-based pixel selection. Moreover, the proposed approach is designed with the amalgamation of 8-pixel stack extraction on the third Level DWT over the Red and Blue channels. Experimental results show that the proposed scheme offers high imperceptibility and enhance the robustness against MPEG-4 compression.

Understanding Heterogeneous EO Datasets: A Framework for Semantic Representations

Earth observation (EO) has become a valuable source of comprehensive, reliable, and persistent information for a wide number of applications. However, dealing with the complexity of land cover is sometimes difficult, as the variety of EO sensors reflects in the multitude of details recorded in several types of image data. Their properties dictate the category and nature of the perceptible land structures. The data heterogeneity hampers proper understanding, preventing the definition of universal procedures for content exploitation. The main shortcomings are due to the different human and sensor perception on objects, as well as to the lack of coincidence between visual elements and similarities obtained by computation. In order to bridge these sensory and semantic gaps, the paper presents a compound framework for EO image information extraction. The proposed approach acts like a common ground between the user’s understanding, who is visually shortsighted to the visible domain, and the machines numerical interpretation of a much wider information. A hierarchical data representation is considered. At first, basic elements are automatically computed. Then, users can enforce their judgement on the data processing results until semantic structures are revealed. This procedure completes a user-machine knowledge transfer. The interaction is formalized as a dialogue, where communication is determined by a set of parameters guiding the computational process at each level of representation. The purpose is to maintain the data-driven observable connected to the level of semantics and to human awareness. The proposed concept offers flexibility and interoperability to users, allowing them to generate those results that best fit their application scenario. The experiments performed on different satellite images demonstrate the ability to increase the performances in case of semantic annotation by adjusting a set of parameters to the particularities of the analyzed data.

Analysis of Multiple Data Hiding Combined Coloured Visual Cryptography and LSB

Currently the level of data security becoming a major factor in data transfer. As we know every process of sending data through any medium the risk of that data gets hacked will still be there. Some techniques for securing data such as steganography and cryptography also often used as a solution for securing data. But it does not last long because it has been found out the weaknesses of the algorithm so that the security be assured. So, in need of variety of new algorithms to be able to protect the data so that data security can be guaranteed. In this study tries to combine two visual algorithms that steganography and cryptography. Where in these experiments will try to secure two pieces of data type that is the type of image data and text data where both the data is regarded as a message so to obtain the correct information receiver should get that two types of data.

Tracking larynx movement in real-time MRI data

Ejective and implosive consonants involve rapid raising or lowering, respectively, of the larynx coordinated with an oral constriction formation and release. Techniques such as contour tracking and region-of-interest (ROI) direct image analysis are becoming established for examining supralaryngeal constriction formation in real-time MRI data, but quantification of larynx movement from dynamic vocal-tract magnetic resonance imaging has not been tackled. We will evaluate methods of indexing the movement of the larynx in fast real-time MRI data collected in the midsagittal plane. We analyze data from USC’s real-time MRI IPA database produced by several phoneticians and data from speakers of the Hausa language producing both ejectives and implosives. Kinematic profiles of the laryngeal movements are obtained by outlining the arytenoid structure and applying principal component analysis on the vocal-tract outlines. In addition, vertical movement of the larynx is indexed by calculating the time-varying pixel intensity centroid (i.e., intensity-weighted average spatial position) in a rectangular ROI defined for larynx movement. Once direct measurement of the larynx actions is validated for this type of imaging data, further work will focus on the temporal coordination of the laryngeal raising/lowering gestures and the supralaryngeal constriction gestures that occur in glottalic airstream consonants. [Work supported by NIH.]

A Subpixel Target Detection Approach to Hyperspectral Image Classification

Hyperspectral image classification faces various levels of difficulty due to the use of different types of hyperspectral image data. Recently, spectral–spatial approaches have been developed by jointly taking care of spectral and spatial information. This paper presents a completely different approach from a subpixel target detection view point. It implements four stage processes, a preprocessing stage, which uses band selection (BS) and nonlinear band expansion, referred to as BS-then-nonlinear expansion (BSNE), a detection stage, which implements constrained energy minimization (CEM) to produce subpixel target maps, and an iterative stage, which develops an iterative CEM (ICEM) by applying Gaussian filters to capture spatial information, and then feeding the Gaussian-filtered CEM-detection maps back to BSNE band images to reprocess CEM in an iterative manner. Finally, in the last stage Otsu’s method is applied to converting ICEM-detected real-valued maps to discrete values for classification. The entire process is called BSNE-ICEM. Experimental results demonstrate BSNE-ICEM, which has advantages over support vector machine-based approaches in many aspects, such as easy implementation, fewer parameters to be used, and better false classification and precision rates.

JUVIL: A new innovative software framework for data analysis of JET imaging systems intended for the study of plasma physics and machine operational safety

A new powerful software framework JUVIL (JET Users Video Imaging Library) has been developed and successfully installed at JET for fast data visualization and advanced analysis of all types of imaging data. The JUVIL framework is based on modular object-oriented components implemented in Python to simplify work with JET scientific data. It provides standard interfaces to access video data and post-processing, which are highly configurable and can be easily extended and adapted for new data formats and imaging cameras. One of the GUI components is the video player, widely used during the last JET campaign. It displays the video data for NIR/IR/VIS cameras and automatically carries out the post-processing (image rotation, data format conversion, scaling of non-interlaced fields to full frames).

A New Method Based on Two-Stage Detection Mechanism for Detecting Ships in High-Resolution SAR Images

Ship detection in synthetic aperture radar (SAR) remote sensing images, being a fundamental but challenging problem in the field of satellite image analysis, plays an important role for a wide range of applications and is receiving significant attention in recent years. Aiming at the requirements of ship detection in high-resolution SAR images, the accuracy, the intelligent level, a better real-time operation and processing efficiency, The characteristics of ocean background and ship target in high-resolution SAR images were analyzed, we put forward a ship detection algorithm in high-resolution SAR images. The algorithm consists of two detection stages: The first step designs a pre-training classifier based on improved spectral residual visual model to obtain the visual salient regions containing ship targets quickly, then achieve the purpose of probably detection of ships. In the second stage, considering the Bayesian theory of binary hypothesis detection, a local maximum posterior probability (MAP) classifier is designed for the classification of pixels. After the parameter estimation and judgment criterion, the classification of pixels are carried out in the target areas to achieve the classification of two types of pixels in the salient regions. In the paper, several types of satellite image data, such as TerraSAR-X (TS-X), Radarsat-2, are used to evaluate the performance of detection methods. Comparing with classical CFAR detection algorithms, experimental results show that the algorithm can achieve a better effect of suppressing false alarms, which caused by the speckle noise and ocean clutter background inhomogeneity. At the same time, the detection speed is increased by 25% to 45%.