Abstract: The present disclosure describes a system and method to segment optical coherence tomography (OCT) images. The present system uses a hybrid method that employs both Bayesian level sets (BLS) and graph-based segmentation algorithms. The system first identifies retinal tissue within an OCT image using the BLS algorithms. The identified retinal tissue is then further segmented using the graph-based segmentation algorithms.

Abstract: A tracking apparatus includes a tracking subject setting unit, a tracking subject search unit, an obstructing detection unit, and an obstructing area detection unit. The tracking subject search unit searches for the tracking subject by using at least one of luminance of the image data, color of the image data, a result of face detection, and a result of focus detection. The obstructing detection unit detects that the tracking subject is obstructed by comparing a standard focus detection information with a focus detection information obtained at a tracking position. The obstructing area detection unit detects an obstructing area. The tracking subject search unit exclusively sets a search area around the obstructing area and searches the set search area for the tracking subject by using information different from the result of focus detection, when the obstructing of the tracking subject is detected.

Abstract: A computer-implemented method and an apparatus are provided. The method includes obtaining, by a processor, a frequency spectrum of an audio signal data. The method further includes extracting, by the processor, periodic indications from the frequency spectrum. The method also includes inputting, by the processor, the periodic indications and components of the frequency spectrum into a neural network. The method additionally includes estimating, by the processor, sound identification information from the neural network.

Abstract: An image recognition device includes an identification unit configured to compare a feature point in an input image and a feature point in every model image to compute a first degree of similarity between the input image and the model image, and to identify the input image on the basis of said first degree of similarity. To compute the first degree of similarity for a model image, the identification unit adds to a score based on a matching feature point for each feature point in the model image matching a feature point in the input image. The score based on the matching feature point is a value that increases as the number of model images including the matching feature point decreases.

Abstract: The disclosure relates to various distance metrics that may quantify semantic and syntactic relationships between devices. More particularly, a first grammar associated with a first device and a second grammar associated with a second device may each comprise a symbol sequence that re-expresses one or more sequenced data items and one or more rules that represent a repeated pattern in the symbol sequence. Accordingly, one or more distance metrics that quantify a similarity between the first grammar and the second grammar may be calculated according to a comparison between the rules in the first grammar and the rules in the second grammar such that a relationship between the first device and the second device can be determined according to the one or more distance metrics.

Abstract: A method including receiving, by an image classification engine, a number K of clusters to be created by a clustering algorithm. The method further including receiving, by the image classification engine, a set of elements based on an image, executing, by the image classification engine, the clustering algorithm on the set of elements to create K clusters, each cluster having a respective subset of the set of elements, for each cluster, computing, by the image classification engine, a centroid of the cluster, for each cluster, creating, by the image classification engine, a generator of the cluster based on the respective subset of the set of elements corresponding to each cluster, and for each element of each cluster, computing, by the image classification engine, a cost function corresponding to the element based on the centroid, the respective subset of the set of elements, and the generator corresponding to the cluster.

Abstract: The disclosure provides a method, electronic device, and computer-readable storage medium for monitoring breathing activity. The method includes the following steps. A time series of incoming frames of a monitored subject is received and each incoming frame is spitted into multiple blocks, where each block of the time series of incoming frames with a same position forms a time series of incoming blocks. A motion vector of each block is computed and projected onto multiple non-parallel axes to be assigned a sign. The projected motion vectors of each time series of incoming blocks with respect to each axis are respectively merged into a corresponding 1D waveform and converted into a frequency domain, and an observe area is located accordingly. A breathing frequency of the monitored subject corresponding to the observe area is determined and whether or not the monitored subject is breathing abnormally is thereby determined.

Abstract: Provided is an image processing apparatus including a memory that stores information, and a reduced image generation unit that extracts a row including a distinguishable part from other rows as a heading from a vertically long range from top to bottom on a left end side when document data is horizontally written, extracts a row including a distinguishable part from other rows as a heading from a horizontally long range from a right end to a left end on an upper end side when the document data is vertically written, and generates a portion including the heading as a reduced image.

Abstract: An electronic device able to receive a target frame for tracking a moving target which is included in a search reference sets at least one first detecting region in the target frame. The processor determines similarities between the search reference and the at least one first detecting regions. Then, the processor determines a specific first detecting region in the target frame and determines an actual position of the moving target based on the specific first detecting region.

Abstract: To measure a three-dimensional shape of a measurement target object at high speeds, an information processing apparatus is configured to acquire an image captured while a pattern including first features and a second feature is projected onto the measurement target object; identify the first features in the captured image, and acquire an initial value of shape information on the measurement target object in an area including the first features; set a candidate for the shape information in a first area based on an initial value of shape information in an area surrounding the first area; decide an area in the pattern corresponding to the first area as a second area, and evaluate a correlation between the second area and the first area based on the second feature in the first area and the second feature in the second area; and derive the shape information in the first area.

Abstract: A system and method, which enables precise identification of characters contained in vehicle license plates, container I.D, chassis I.D, aircraft serial number and other such identification markings. The system can process these identified characters and operate devices, such as access control operations, traffic systems and vehicle and container tracking and management systems, and provide records of all markings together with their images.

Abstract: Described is a training system for training a vision-based object detector. The system is configured to run an object detector on an image of a cleared scene to detect objects in the cleared scene. The object detector includes a support vector machine (SVM) or similar classifier with a feature model to generate an SVM score for object features and a spatial bias threshold to generate augmented object scores. The system designated detected objects in the cleared scene as false detections and, based on that, updates at least one of the feature model and spatial bias threshold to designate the false detections as background. The updated feature model or updated spatial bias threshold are then stored for use in object detection.

Abstract: One of the challenges in bringing computational imaging to a mass market is that computational imaging is inherently computationally expensive. The computational challenges associated with computational imaging are apparent with the computation of a histogram of gradient descriptors. Oftentimes, generating a histogram of gradient descriptors involves computing gradients of an image, binning the gradients according to their orientation, and, optionally, normalizing the bins using a non-linear function. Because each of these operations is expensive, the histogram of gradient descriptor computations is generally computationally expensive and is difficult to implement in a power efficient manner for mobile applications. The present application discloses a computing device that can provide a low-power, highly capable computing platform for computing a histogram of gradient descriptors.

Abstract: Systems and methods for tracking telecom computing device vulnerabilities. The system includes a database storing a plurality of entries that describe security vulnerabilities, and a controller that receives input from a user selecting a class of telecommunication devices, e.g. a product line, and searches the database to identify pertinent entries describing a relevant security vulnerability for the class. The controller also identifies an authoritative entry that describes the relevant security vulnerability, validates pertinent entries within the database that conform with the authoritative entry, generates a report indicating a severity of the security vulnerability within the class based on valid entries within the database, and assesses a severity of the relevant security vulnerability for the class based on the conforming entries.

Abstract: A system comprises a picture and metadata captured by a content capture system; a recognizable characteristic datastore configured to store recognizable characteristics of different users; a module configured to identify a time and a location associated with the picture based on the metadata, and to identify one or more potential target systems within a predetermined range of the location at the time; a characteristic recognition module configured to retrieve the recognizable characteristics of one or more potential users associated with the potential target systems, and evaluate whether the picture includes one or more representations of at least one actual target user from the potential users based on the recognizable characteristics of the potential users; a distortion module configured to distort a feature of the representations of the least one actual target user in response to the determination; a communication module configured to communicate the distorted picture to a computer network.

Abstract: An operation estimation apparatus includes a distance measurement unit configured to measure a distance for a predetermined range, a human-body processing unit configured to detect a person on the basis of a distribution of measured distances, a tracking unit configured to assign an identifier to each detected person, and a situation estimation unit configured to estimate reaction of the person on the basis of a human data height which is a value of a height direction included in the measured distance in a first period set by prompting the person to start and end hand raising, wherein the situation estimation unit measures the human data height for each person to which the identifier is assigned a plurality of times in the first period, extracts an identifier for which human data heights are in the predetermined range, and estimates the reaction of the person corresponding to the extracted identifier.

Abstract: An object is to ensure clear and easy quantification of the textures such as metallic texture and shiny texture of pearl pigment and to rationalize comparison inspection between an inspection object and a reference object. A coloring inspection apparatus 1 includes a camera 2 that is configured to have three spectral sensitivities (S1(?), S2(?), S3(?)) linearly and equivalently converted to a CIE XYZ color matching function, an arithmetic processing unit 3 that is configured to obtain and compute coloring data by conversion of an image which has three spectral sensitivities and is obtained by the camera 2 into tristimulus values X, Y and Z in a CIE XYZ color system, and lighting units 6 that are configured to illuminate an automobile 5 as an example of measuring object. The coloring inspection apparatus 1 computes a color distribution consistency index that represents a ratio of overlap of two xyz chromaticity histogram distributions of an inspection object Q and a reference object R, so as to inspect color.

Abstract: A learning method for improving performance of a CNN by using Feature Up-sampling Networks is disclosed. The learning method includes steps of: (a) allowing the down-sampling block to acquire a down-sampling image; (b) allowing each of a (1-1)-th to a (1-k)-th filter blocks to respectively acquire each of a (1-1)-th to a (1-k)-th feature maps; (c) allowing a specific up-sampling block to (i) receive a particular feature map from its corresponding filter block, and (ii) receive another specific feature map from its previous up-sampling block, and then rescale a size of the specific feature map to be identical to that of the particular feature map and (iii) apply certain operations to the particular feature map and the resealed specific feature map to generate a feature map of the specific up-sampling block; and (d)(i) allowing an application block to acquire an application-specific output and (ii) performing a first backpropagation process.

Abstract: A system automatically and continuously finds and aggregates the most relevant and current information about the people and things that a user cares about. The information gathering is based on current context (e.g., where the user is, what the user is doing, what the user is saying/typing, etc.). The result of the context based information gathering is presented ubiquitously on user interfaces of any of the various physical devices operated by the user.

Abstract: A computing device determines a contrast medium uptake time using magnetic resonance imaging data. Image data constructed from data generated by a magnetic resonance imaging (MRI) machine of a subject is read. A representation computed from the read image data is presented on a display device. Baseline artery locations identified within the presented representation that are associated with a baseline artery are received. A first time-of-arrival (TOA) of contrast medium into the baseline artery is determined using the received baseline artery locations and the read image data. For a plurality of locations within the read image data excluding the baseline artery locations, a second TOA of the contrast medium into a respective location relative to the determined first TOA is determined using the read image data, and the determined second TOA is stored in association with the respective location to assist in lesion identification for the subject.

Abstract: A picture processing method and apparatus are described. The method includes: detecting newly-added pictures in a photo album in a computing device; acquiring information about the newly-added pictures, the information including either or both of a photographing time and a photographing place; updating a to-be-processed picture set according to the information about the newly-added pictures; selecting a picture processing template matching the to-be-processed picture set; and processing pictures in the to-be-processed picture set by using the picture processing template. The method can automatically process a newly-added picture, thereby reducing time costs of a user for picture processing.

Abstract: An image processing system may include an imaging device for capturing an image and an image processing apparatus for processing the image. The imaging device may include an imaging unit for capturing the image, a first recording unit for recording information relating to the image, the information being associated with the image, and a first transmission control unit for controlling transmission of the image to the image processing apparatus. The image processing apparatus may include a reception control unit for controlling reception of the image transmitted from the imaging device, a feature extracting unit for extracting a feature of the received image, a second recording unit for recording the feature, extracted from the image, the feature being associated with the image, and a second transmission control unit for controlling transmission of the feature to the imaging device.

Abstract: A terminal is provided that includes a speech acquisition unit that acquires first speech information, and a first speech processor that removes noise contained in the first speech information using a first removal method and outputs the noise-removed speech information as second speech information. The terminal also includes a first speech recognition unit that performs speech recognition on the second speech information and outputs the speech recognition result as first speech recognition result information. The terminal further includes a communication unit that receives a speech recognition result as second speech recognition result information from a server, and a determination unit that makes a selection as to which of the first speech recognition result information and the second speech recognition result information should be outputted.

Abstract: A moving object controller efficiently generates an environmental map and performs highly accurate state estimation in a short time to appropriately control a moving object. An observation obtaining unit obtains observation data from an observable event. A landmark prediction unit generates a landmark prediction signal including predictive information about a landmark at a current time. A landmark detection unit detects information about the landmark at the current time, and generates a landmark detection signal indicating the detection result. A state estimation unit estimates an internal state of the moving object to obtain data indicating an estimated internal state of the moving object at the current time, and estimates the environmental map based on the landmark detection signal to obtain data indicating an estimated environmental map at the current time.

Abstract: An illuminating device coupled with sensors or an image acquisition device and a logical controller allows illumination intensity and spectrum to be varied according to changing user needs. The system provides illumination to areas according to the principles of correct lighting practice for the optimal performance of visual tasks in the most efficient, cost effective manner. Aspects of the invention include: lighting fixtures which adapt to ambient lighting, movement, visual tasks being performed, and environmental and personal conditions affecting illumination requirements at any given instant. Lighting fixtures having spatial distribution of spectrum and intensity, providing both “background” room lighting, and “task” lighting.

Abstract: The video processing system provides for automatic selection of video sequences showing goals in a video stream from cameras directed at a playing field. Classification is applied to image content from different foreground positions in an image. The set of classes comprising a first and second class corresponding players of different teams. Medians or averages of the image positions of the local image content assigned to the first and second class are determined. A condition is tested whether the results are on mutually opposite sides of a reference position and a distance between collected positions or distances between the reference position and the first and second collected positions exceed a threshold value. When the test shows that the condition is met, a preceding video sequence is selected, for example for display.

Abstract: To display a news feed in a social network environment, a social networking system generates news items regarding activities associated with a user of a social network environment. The social networking system may also attach an informational link associated with at least one of the activities to at least one of the news items, limit access to the news items to a predetermined set of viewers, and assign an order to the news items. The news items may be displayed in the assigned order to at least one viewing user of the predetermined set of viewers, and the number of news items displayed may be dynamically limited.

Abstract: A method, apparatus and computer program product are hereby provided to detect anatomical elements in a medical image. In this regard, the method, apparatus, and computer program product may receive a test image and generate a classified image by applying an image classifier to the test image. The image classifier may include at least one decision tree for evaluating at least one pixel value of the test image and the classified image may include a plurality of pixel values. Each pixel value may be associated with a probability that an anatomical element is located at the pixel location. The method, apparatus, and computer program product may also evaluate the classified image using an anatomical model to detect at least one anatomical element within the classified image.

Abstract: A computer detects one or more images, wherein detecting the one or more images further includes identifying one or more characteristics of at least a portion of the one or more images. The computer detects one or more user communications. The computer determines one or more appropriate images of the one or more images to be associated with the one or more user communications based on comparing information associated with the one or more user communications with the identified one or more characteristics of at least a portion of the one or more images.

Abstract: The present embodiments relate to machine learning for multimodal image data. By way of introduction, the present embodiments described below include apparatuses and methods for learning a similarity metric using deep learning based techniques for multimodal medical images. A novel similarity metric for multi-modal images is provided using the corresponding states of pairs of image patches to generate a classification setting for each pair. The classification settings are used to train a deep neural network via supervised learning. A multi-modal stacked denoising auto encoder (SDAE) is used to pre-train the neural network. A continuous and smooth similarity metric is constructed based on the output of the neural network before activation in the last layer. The trained similarity metric may be used to improve the results of image fusion.

Abstract: An image processing apparatus generates one image by using at least one frame each of a plurality of moving images obtained by taking moving images of a plurality of different regions of an eye at different times. The apparatus includes a deciding unit configured to decide the at least one frame in each of the plurality of moving images, so that regions which have actually been shot are included in the plurality of moving images in the plurality of regions; and an image generating unit configured to generate one image by using the at least one frames decided from each of the plurality of moving images.

Abstract: The present invention is to provide a system, a method, and a program for identifying a person depicted in a portrait. The system for identifying a person depicted in a portrait 1 receives input of portrait data of the portrait, stores different individual data depending on a plurality of persons, extracts a feature point of the received portrait by image analysis, checks the extracted feature point against the individual data, identifies a person depicted in the portrait from the check result, and displays the identification result together with a probability corresponding to an identical level.

Abstract: A system for performing global scale object detection using satellite imagery, comprising an object detection server that receives and analyzes image data to identify objects within an image via a curated computational method, and a curation interface that enables a user to curate image information for use in object identification, and a method for a curated computational method for performing global scale object detection.

Abstract: In one example, a method includes sending, from a mobile device and to a remote server, a message indicating a first captured image corresponds to a first sample image in response to determining that the degree of similarity between the first captured image and the first sample image is within a first similarity interval. The method includes sending, to the remote server, a message representing at least part of a second captured image in response to determining that the degree of similarity between the second captured image and a second sample image is within a second similarity interval. The method also includes refraining from sending a message associated with a third captured image to the server in response to determining that the degree of similarity between the third captured image and a third sample image is within the first or second similarity intervals.

Abstract: A system and method of for describing target data as a sequence of pattern elements and pattern element groups that comprise an overall target pattern is described. Pattern elements may utilize regular expression syntax along with other metadata that describe the behavior of the element. A pattern element group may be a collection of fully defined pattern elements where at least one pattern element from the group must have a match for the overall pattern to match. Patterns contain both pattern elements and pattern element groups. The general process involves first performing optical character recognition (OCR) on the document, which in turn produces a sequence of text tokens representing the lines of text on each page of the document. The search algorithm may then apply each defined pattern to the entire document capturing and/or extracting data that match each pattern's required elements and element groups.

Abstract: The present disclosure relates to image retrieval in an electronic device. An operating method of an electronic device includes determining an object included in an image and a position of extent occupied by the object in the image and retrieving at least one image including the object at the position from among a plurality of images.

Abstract: Described is a technology by which an image such as a stitched panorama is automatically cropped based upon predicted quality data with respect to filling missing pixels. The image may be completed, including by completing only those missing pixels that remain after cropping. Predicting quality data may be based on using restricted search spaces corresponding to the missing pixels. The crop is computed based upon the quality data, in which the crop is biased towards including original pixels and excluding predicted low quality pixels. Missing pixels are completed by using restricted search spaces to find replacement values for the missing pixels, and may use histogram matching for texture synthesis.

Abstract: A processor-implemented method and system of this disclosure are configured to correct or resolve artifacts in a received digital image, using a selection of one or more encompassment measures, each encompassing the largest artifact and an optional second smaller artifact. The processor herein is configured to calculate difference in Gaussians using blurred versions of the input digital image and optionally, the input digital image itself; to composite the resulting difference in Gaussians and the input digital image; and to determine pixels with properties of invariant values, also referred to as invariant pixels. The values in the properties of the invariant pixels are then applied to the artifacts regions to correct these artifacts, thereby generating the modified digital image in which the artifact regions are more or less differentiable to a human eye when compared with the digital image.

Abstract: An image analysis unit of an image processing apparatus acquires a distribution condition of feature points in an entire input image and in each of a plurality of small regions in the input image. A target point number setting unit sets a target point number for each small region based on the distribution condition of feature points in the entire input image. An expected point number setting unit sets an expected point number for each small region based on the distribution condition of feature points in the small region. A comparison unit compares the target point number and the expected point number. A sensitivity setting unit sets detection sensitivity based on the comparison result. A feature point detection unit performs feature point detection according to the detection sensitivity.

Abstract: An image recognition apparatus includes a storage unit which stores a plurality of pieces of identical recognition reference data which are used as references for recognizing a specific object from an image, an image recognition unit which recognizes the specific object by alternately making a switch between the plurality of pieces of recognition reference data, on a frame basis, stored in the storage unit, for consecutive frames of the image acquired by an image acquisition unit, and a recognition determining unit which, on condition that recognition results obtained by the image recognition unit by using the plurality of pieces of recognition reference data meet a predetermined matching condition, outputs the recognition results as effective recognition results.

Abstract: A data processing method is disclosed for processing hyperspectral image data of a scene. The method comprises sequentially receiving portions of the data at a data buffer to form a data set comprising a predefined number of data portions and calculating a set of global statistical parameters and data correction factors using the data forming the data set. The method further comprises receiving a further data portion at the data buffer and simultaneously removing the earliest received data portion at the data buffer, from the data set, and subsequently calculating a set of local statistical parameters using the data of the further data portion. The method further comprises updating the set of global statistical parameters using the set of local statistical parameters and correcting the data of the data portion removed from the data set using the correction factors.

Abstract: Recommendations can be generated even in situations where sufficient user information is unavailable for providing personalized recommendations. Instead of generating recommendations for an item based on item type or category, a relation graph can be consulted that enables other items to be recommended that are related to the item in some way, which may be independent of the type or category of item. For example, images of models, celebrities, or everyday people wearing items of clothing, jewelry, handbags, shoes, and other such items can be received and analyzed to recognize those items and cause them to be linked in the relation graph. When generating recommendations or selecting advertisements, the relation graph can be consulted to recommend products that other people have obtained with the item from any of a number of sources, such that the recommendations may be more valuable to the user.

Abstract: An image processing device includes a dithering circuit, a gradation correction pattern generating circuit, and a gradation correction pattern correcting circuit. The dithering circuit performs a dithering process on image data to generate dithered image data. The gradation correction pattern generating circuit sets, in multi-valued data, gradation information of each of pixels of the generated dithered image data, and generates a gradation correction pattern imitating the dithered image data. The gradation correction pattern correcting circuit performs pattern matching on the generated gradation correction pattern, and corrects the generated gradation correction pattern in accordance with a result of the pattern matching.

Abstract: Systems, methods and devices for automatically cropping images taken by an electronic device in order to determine the identity of a product contained in the image are described herein. A number of different techniques may be applied to perform the automatic cropping, including a focus sweep technique in which a first image is analyzed for the presence of a human being and then a second image is taken in a plane closer to the camera than the first image. The two frames are analyzed and a resultant image is provided that avoids the regions in which the human being is present to focus on the product. In other embodiments, a motion vector calculation is made between two images in which an individual is holding a product. The motion vectors related to the human are removed and a bounding box is calculated to reduce the size of the image to include a higher percentage of the product, such that the product can be more easily identified.

Abstract: A method and a system for summarizing a concept are provided. A query corresponding to a concept is received from a user. A plurality of images and corresponding descriptive information may be collected based on the query. The plurality of images and the descriptive information may be processed to form feature vectors and processed descriptive information respectively. Further, one or more topics may be identified for the plurality of images. Each of the plurality of images may be assigned with one or more topic distribution values corresponding to the one or more topics. The one or more topics correspond to the processed descriptive information. A sparse set of images may be determined based on the feature vectors and the assigned topic distribution values, to summarize the concept. Also, a target summary may be built from the summarized concept, by regularizing one or more distribution constraints.

Abstract: Methods, systems, computer-readable media, and apparatuses for incremental object detection using a staged process and a band-pass feature extractor are presented. At each stage of the staged process, a different band of features from a plurality of bands of features in image data can be extracted using a dual-threshold local binary pattern operator, and compared with features of a target object within the band for a partial decision. The staged process exits if a rejection decision is made at any stage of the staged process. If no rejection decision is made in each stage of the staged process, the target object is detected. Features extracted at each stage may be from a different image for some applications.

Abstract: Various embodiments provide a method for determining conditions associated with the presence of a user in order to perform object detection. For example, various heuristic tests can be used to identify movement within a sequence of images or to identify user engagement therewith and, if movement or user engagement is identified, stereo disparity can be computed in an attempt to detect an object within a predetermined distance of the computing device. Accordingly, if the distance of the object is within the predetermined distance, object detection is performed. If the object is beyond the threshold distance, however, object detection is not performed in order to conserve battery life of the computing device.

Abstract: A system for performing global-scale damage detection using satellite imagery, comprising a damage detection server that receives and analyzes image data to identify objects within an image via a curated computational method, and a curation interface that enables a user to curate image information for use in object identification, and a method for a curated computational method for performing global scale damage detection.

Abstract: A system and method for generating automated response to an input query received from a user in a human-machine interaction environment are described. The system may comprise an external memory wherein the data is stored and segregated into a plurality of segments in the hierarchical structure. The system may further comprise a processor and a memory coupled with the processor. The processor may execute a plurality of modules stored in the memory. A segment identification module may be configured to identify a relevant segment, from the plurality of segments, matching with a user input. A relevant data determination module may be configured to determine relevant data within the relevant segment matching with the user input. A response generation module may be configured to generate a response with respect to the user input based upon the relevant data.

Abstract: Embodiments are disclosed that relate to interacting with a graphical user interface in a manner to facilitate hands-free operation. For example, one disclosed embodiment provides a method including outputting to a display device a user interface that displays graphical representations of a plurality of alignable user interface objects, each alignable user interface object representing a selectable object. The method further includes receiving depth data from an imaging device, the image data capturing an image of a face directed toward the display device. The method further comprises changing an alignment condition of a first user interface object of the plurality of alignable user interface objects to move the first user interface object into an aligned condition, and changing an alignment condition of a second user interface object to move the second user interface object out of an aligned condition based on the received image data.