Showing papers on "Image conversion published in 2012"

Semi-coupled dictionary learning with applications to image super-resolution and photo-sketch synthesis

Abstract: In various computer vision applications, often we need to convert an image in one style into another style for better visualization, interpretation and recognition; for examples, up-convert a low resolution image to a high resolution one, and convert a face sketch into a photo for matching, etc. A semi-coupled dictionary learning (SCDL) model is proposed in this paper to solve such cross-style image synthesis problems. Under SCDL, a pair of dictionaries and a mapping function will be simultaneously learned. The dictionary pair can well characterize the structural domains of the two styles of images, while the mapping function can reveal the intrinsic relationship between the two styles' domains. In SCDL, the two dictionaries will not be fully coupled, and hence much flexibility can be given to the mapping function for an accurate conversion across styles. Moreover, clustering and image nonlocal redundancy are introduced to enhance the robustness of SCDL. The proposed SCDL model is applied to image super-resolution and photo-sketch synthesis, and the experimental results validated its generality and effectiveness in cross-style image synthesis.

2D-to-3D image conversion by learning depth from examples

Abstract: Among 2D-to-3D image conversion methods, those involving human operators have been most successful but also time-consuming and costly. Automatic methods, that typically make use of a deterministic 3D scene model, have not yet achieved the same level of quality as they often rely on assumptions that are easily violated in practice. In this paper, we adopt the radically different approach of “learning” the 3D scene structure. We develop a simplified and computationally-efficient version of our recent 2D-to-3D image conversion algorithm. Given a repository of 3D images, either as stereopairs or image+depth pairs, we find k pairs whose photometric content most closely matches that of a 2D query to be converted. Then, we fuse the k corresponding depth fields and align the fused depth with the 2D query. Unlike in our original work, we validate the simplified algorithm quantitatively on a Kinect-captured image+depth dataset against the Make3D algorithm. While far from perfect, the presented results demonstrate that online repositories of 3D content can be used for effective 2D-to-3D image conversion.

Automatic 2D-to-3D image conversion using 3D examples from the internet

Abstract: The availability of 3D hardware has so far outpaced the production of 3D content. Although to date many methods have been proposed to convert 2D images to 3D stereopairs, the most successful ones involve human operators and, therefore, are time-consuming and costly, while the fully-automatic ones have not yet achieved the same level of quality. This subpar performance is due to the fact that automatic methods usually rely on assumptions about the captured 3D scene that are often violated in practice. In this paper, we explore a radically different approach inspired by our work on saliency detection in images. Instead of relying on a deterministic scene model for the input 2D image, we propose to "learn" the model from a large dictionary of stereopairs, such as YouTube 3D. Our new approach is built upon a key observation and an assumption. The key observation is that among millions of stereopairs available on-line, there likely exist many stereopairs whose 3D content matches that of the 2D input (query). We assume that two stereopairs whose left images are photometrically similar are likely to have similar disparity fields. Our approach first finds a number of on-line stereopairs whose left image is a close photometric match to the 2D query and then extracts depth information from these stereopairs. Since disparities for the selected stereopairs differ due to differences in underlying image content, level of noise, distortions, etc., we combine them by using the median. We apply the resulting median disparity field to the 2D query to obtain the corresponding right image, while handling occlusions and newly-exposed areas in the usual way. We have applied our method in two scenarios. First, we used YouTube 3D videos in search of the most similar frames. Then, we repeated the experiments on a small, but carefully-selected, dictionary of stereopairs closely matching the query. This, to a degree, emulates the results one would expect from the use of an extremely large 3D repository. While far from perfect, the presented results demonstrate that on-line repositories of 3D content can be used for effective 2D-to-3D image conversion. With the continuously increasing amount of 3D data on-line and with the rapidly growing computing power in the cloud, the proposed framework seems a promising alternative to operator-assisted 2D-to-3D conversion.

Image display device, image display method, and program

Abstract: An image display device includes region of interest extraction unit, parallax image generation unit and 3D image display unit. Region of interest extraction unit generates depth image signal by depth image conversion employing depth threshold, depth image signal including information on distance in three-dimensional space between viewpoint and each pixel of two-dimensional image including region of interest desired to be noted by observer, depth image conversion being such that depth value indicating distance between viewpoint and each pixel of two-dimensional image is converted to depth value for 2D display when depth value is equal to or larger than depth threshold. Parallax image generation unit generates both-eye parallax image having parallax in region of interest alone, from two-dimensional image and image obtained by conversion of region of interest image representing region of interest at each of both-eye viewpoints, based on two-dimensional image and depth image signal. 3D image display unit displays both eye parallax image.

Depth image conversion apparatus and method

Abstract: Provided are an apparatus and method for converting a low-resolution depth image to a depth image having a resolution identical to a resolution of a high-resolution color image. The depth image conversion apparatus may generate a discrete depth image by quantizing a depth value of an up-sampled depth image, estimate a high-resolution discrete depth image by optimizing an objective functions of the discrete depth image based on the high-resolution color image and an up-sampled depth border, and convert the up-sampled depth image to a high-resolution depth image by filtering the up-sampled depth image when a difference between discrete depth values of neighboring pixels in the high-resolution discrete depth image is less than a predetermined threshold value.

Method and device for synthesizing virtual viewpoints in multi-viewpoint video

Abstract: The invention discloses a method and device for synthesizing virtual viewpoints in a multi-viewpoint video, which are used for improving a quality and a display effect of a virtual viewpoint image on the premise of guaranteeing that a geometric distortion does not occur. The method comprises the following steps of: using two viewpoint images provided by a first camera and a second camera as reference viewpoint images, and carrying out three-dimensional image conversion, small cavity filling, image complementation processing and image fusion processing on the two viewpoint images, respectively to obtain a main virtual viewpoint image; and carrying out depth map processing, the three-dimensional image conversion, the small cavity filling, the image complementation processing and the image fusion processing on the two viewpoint images, respectively to obtain a secondary virtual viewpoint image; and filling residual cavities of the main virtual viewpoint image according to the secondary virtual viewpoint image to obtain a virtual viewpoint image in need of being synthesized finally.

Video communication system and video communication method

Abstract: PROBLEM TO BE SOLVED: To provide a video communication system, capable of making an expression of an interactive opposition party in a video more realistic to a user.SOLUTION: A video communication system includes: image acquisition means for photographing a two-dimensional image involving an imaging side person and a space other than the person; distance image acquisition means for acquiring a distance image involving the imaging side person and the space other than the person; three-dimensional model production means for producing three-dimensional model data for an imaging side space corresponding to the two-dimensional image on the basis of the distance image; three-dimensional image production means for producing a three-dimensional image of the imaging side space by pasting the two-dimensional image as a texture onto the surface of the three-dimensional model defined by the three-dimensional model data acquired from the three-dimensional model production means; viewpoint detection means for detecting a viewpoint position of a displaying side person; image conversion means for converting the three-dimensional image to the two-dimensional image subject to the viewpoint position; and display means for displaying the two-dimensional image converted by the display side display device.

Image processing device, image processing method, and program

Abstract: An image processing device including an image input unit that inputs a two-dimensional image signal; a depth information output unit that inputs or generates depth information; a depth information reliability output unit that inputs or generates the reliability of depth information that the depth information output unit outputs; an image conversion unit that inputs an image signal, depth information, and depth information reliability, and generates and outputs a left eye image and a right eye image for realizing binocular stereoscopic vision; and an image output unit that outputs a left eye image and a right eye image, wherein the image conversion unit has a configuration of performing image generation of at least one of a left eye image and a right eye image and executes a conversion process in which the depth information and the depth information reliability are applied as conversion control data during the image conversion.

Moving body detection device and moving body detection method

Abstract: A moving-body-detecting device is provided with an image capture unit, an image conversion unit, a three-dimensional-object-detecting unit, a turn detecting unit, and a control unit. The image capture unit captures images of the rear side of a vehicle. The image conversion unit performs viewpoint conversion of the captured images into bird's-eye view images. The three-dimensional-object-detecting unit generates differential waveform information based on a differential image of bird's-eye view images taken as different points in time, and detects three-dimensional objects based on the differential waveform information. The control unit suppresses detection of the three-dimensional object based on the differential waveform information, in cases in which turning of the vehicle has been detected as compared with cases in which the vehicle is not turning.

Display system with image conversion mechanism and method of operation thereof

Abstract: PROBLEM TO BE SOLVED: To provide a display system with an efficient and high-performance image conversion mechanism for achieving a three-dimensional image, and a method of operation thereofSOLUTION: The method of operation of the display system includes: receiving a current input image with current input pixels; identifying an input object having input object pixels, based on differences in characteristics of the current input pixels; calculating an object pixel depth for the input object pixels from an image depth map of a current input image, based on a weighted average of depth candidates; and generating a processed image which has a perceived depth and is to be displayed on a device, from the current input image with the image depth map and the object pixel depth

Method for converting input image data into output image data, image conversion unit for converting input image data into output image data, image processing apparatus, display device

Abstract: In a method and device for improving image rendition by contrast and/or sharpness enhancement the sharpness enhancement is made dependent on the local average luminance value A mix is made of spatially enhanced image signals is used, wherein for various signal differing spatial frequencies are boosted The mixing factors for mixing of the boosted signals are dependent on a local average luminance value such that the distribution over frequency bands shifts to higher frequencies and sharper enhancement as the luminance value increases

Lighting control system

Abstract: PROBLEM TO BE SOLVED: To control a light environment in a space by using a simple optical measurement method and incorporating how humans sense brightness, thereby realizing energy saving while also maintaining comfort.SOLUTION: A lighting control system 1 comprises a plurality of luminaires 2 installed in a room space X having an opening A through which daylight is able to enter, an illuminance sensor 3 for measuring illuminance, and a control device 4 capable of changing the brightness of the luminaires 2. The control device 4 includes: image generation means 10 for generating a predicted luminance image in which a luminance state in the room space X is predicted on the basis of the illuminance obtained from the illuminance sensor 3; image conversion means 20 for converting the predicted luminance image to an image to be evaluated having luminance contrast and human adaptation added thereto; determination means 30 for determining using the image to be evaluated whether the lighting state of the room space X is appropriate or not; and illumination adjustment means for changing the brightness of the luminaires 2 when the lighting state is determined to be inappropriate by the determination means 30.

Image capturing device, image capturing device control method, and program

Abstract: An image capturing device includes: a plurality of image capturing units that capture images from different viewpoints; an image analyzing unit that detects characteristic points by image analysis of the images captured by the plurality of image capturing units; a control unit that calculates image conversion parameters for correction of inputting characteristic point information generated by the image analyzing unit to correct deviation in the up and down direction of the plurality of images captured by the plurality of image capturing units; and an image conversion unit that performs a conversion process of the captured images of the plurality of image capturing units by applying the image conversion parameters calculated by the control unit.

Ink use amount evaluation apparatus and method, and inkjet apparatus

Abstract: When image is formed by a liquid ejection head, input image data is subjected to conversion processing including tone conversion using a first look-up table (LUT) and density correction or non-uniformity correction in nozzle units using second LUTs. Ink use amount in image formation is predicted from reduced image data which is generated from the input image data. In this, the calculation is made more efficient by applying a third LUT which is compiled according to requirements from the second LUTs, to the reduced image data, rather than using the second LUTs directly. Since the image conversion processing is carried out which combines the first LUT for tone conversion and the third LUT corresponding to the non-uniformity correction, and the ink amount distribution data is calculated from the reduced image data after the conversion, then it is possible to ascertain an accurate ink amount which reflects the image adjustment conditions.

Image processing device, image processing method, and program

Abstract: PROBLEM TO BE SOLVED: To obtain an image processing device and an image processing method which can appropriately performing image conversion processing which converts a photographic image with a face of a person and the like into an image in a painting tone (painting tone image).SOLUTION: An image processing device comprises: a face detection processing unit 314 which detects a face of a person in image data 208 and determines an area corresponding to the detected face; and an image tone conversion processing unit 302 which performs first image tone conversion processing on image data of the detected area, and also performs second image tone conversion processing different from the first image tone conversion processing on image data of other areas. Thus, by performing image tone conversion suitable for the face on a face section in an image, and performing image tone conversion suitable for a background on sections except the face area, it is possible to realize natural image tone conversion overall.

Image processing device, and image processing method, and program

Abstract: A plurality of viewpoint images are acquired and images for three-dimensional image display are generated without causing a reduction in polarization performance. Provided are: a first polarization means which has a first polarization region and a second polarization region that each allows different polarized light to be transmitted therethrough; a second polarization means which is made up of a third polarization region that allows only the transmitted light of either one of the first polarization region or the second polarization region to be transmitted therethrough, and a total transmission region that allows the transmitted light of the first polarization region and the second polarization region to be transmitted therethrough; an imaging element into which the transmitted light of the second polarization means is input; and an image processing unit which executes signal processing for an output signal of the imaging element. The image processing unit generates parallax information in which subject distance is reflected by applying a first viewpoint image that is based on the transmitted light of the total transmission region of the second polarization means, and a second viewpoint image that is based on the transmitted light of the third polarization region of the second polarization means, and generates a left-eye-purpose image and a right-eye-purpose image for three-dimensional image display by means of image conversion processing of the first viewpoint image for which the parallax information is applied.

Object recognition method and recognition apparatus

Abstract: A target fish-eye image (100) for object recognition is split (120) into regions in accordance with the direction of distortion of the target fish-eye image (100). An object recognition portion (130) performs object recognition by using databases (140) prepared in accordance with the regions respectively. The same process is applied to a plurality of target fish-eye images (110) created by rotation of the target fish-eye image (100). A detected coordinate transformation portion (150) restores the obtained position of an object into the original position by inverse rotation, and outputs a thus obtained result as a detection result. Accordingly, detection accuracy in object recognition can be improved, and the quality of data in databases can be reduced. Rotated object images required for database creation are used for creation of databases for object recognition for the fish-eye image. Thus, it is possible to directly use a distorted fish-eye image for object recognition without projective conversion or image conversion based on corrective operation to the fish-eye image.

Image processing device, image processing method

Abstract: PROBLEM TO BE SOLVED: To provide an image processing device that provides a technology for collecting an image pattern including a recognition target and an image pattern not including a recognition target without time and effort, and using the thus collected image patterns for learning of a discriminator, and thereby carries out highly accurate image recognition.SOLUTION: An image processing device includes: an object tracking part 31 that extracts an area where a recognition target is captured from an image of respective frames that constitute a moving image; an image conversion part 32 that creates a recognition target sample based on an image obtained by subjecting the image in the area to geometric conversion; an area cutting part 34 that sets an area to the image in the frames that constitute the moving image; an image composition part 35 that creates a non-recognition target sample image based on an image obtained by combining plural areas of the image in respective set areas; and a learning part 40 that learns the recognition target using the recognition target sample and the non-recognition target sample.

Object detection device, object detection method, and object detection program

Abstract: An object detection device (10) is equipped with: an image conversion unit (20) that converts an input image produced by photographing the vicinity of a vehicle into a characteristic image for which image characteristics have been extracted from the input image; a per-distance image extraction-combination unit (30) which, on the basis of the distance from the vehicle, extracts from the characteristic image different regions, which differ for each distance, and which combines these regions; a first object detection unit (40) that scans the combined image and performs a first object detection process; an object candidate position designation unit (50) that obtains an object candidate position from the result of the first object detection process; a second object detection unit (60) that performs a second object detection process with respect to the object candidate position; and an object position identification unit (70) that identifies the object position from the result of the second object detection process. Thus, objects in the vicinity of a vehicle can be detected quickly without a decrease in detection capability.

Method and system for enhancing color images of cotton pseudo foreign fibers under non-uniform illumination

Abstract: The invention discloses a method for enhancing color images of cotton pseudo foreign fibers under non-uniform illumination. The method comprises the following steps of: 1, reading original color images of the cotton pseudo foreign fibers in; 2, converting the original color images of the cotton pseudo foreign fibers from an RGB (red, green and blue) space to an HSI (hue, saturation and intensity) space; 3, enhancing components I of the converted HSI color images; 4, performing enhancement post-processing on the images enhanced by the step 3; and 5, converting the HSI images processed by the step 4 back to the RGB space. In addition, the invention also discloses a system for implementing the method. The system comprises an image reading module, a first image conversion module, an image enhancement module, an image enhancement post-processing module and a second image conversion module which are used for realizing the five steps respectively. By utilizing the method and system disclosed by the invention, the color images of the cotton pseudo foreign fibers can be effectively enhanced, so that the subsequent segmentation and identification of foreign fibers and the pseudo foreign fibers can be conveniently realized.

Image conversion apparatus, camera, image conversion method, and storage medium with program stored therein

Abstract: Provided are an image conversion apparatus, a camera, an image conversion method, and a program, wherein a panorama image can be partially enlarged or reduced, while maintaining the characteristics of the panorama image in that images can be joined into a cylindrical shape, or in that a wide range (up to 360 degrees) can be observed simultaneously. This image conversion apparatus adopts a configuration wherein the image conversion apparatus is provided with: an image input unit that inputs a panorama image; and an image enlargement/reduction unit that enlarges/reduces the panorama image by making the enlargement ratio of the desired area within the panorama image different in a way that will make the enlargement ratio change successively at successive areas of the panorama image. In this configuration, the image enlargement/reduction unit enlarges/reduces the image such that the width, in the longitudinal direction of the panorama image, of the image after the enlargement/reduction will fit into the original panorama image.

Method and device for positioning and identifying compressed image object

Abstract: The invention discloses a method and a device for fast searching, positioning and identifying an image object of a JPG image. The traditional method firstly converts the JPG image into a BMP image and then carrying out searching, positioning and identifying on the BMP image. Different from the traditional method, for an image object with obvious visual characteristics, by using the obvious low-frequency information of the image object, the method searches and positions the image object by utilizing frequency domain information when obtaining the image frequency domain information in the process of conversion from the JPG image into a BMP image, and converts an positioned image object area into a partial BMP image, and identifies the image object on the partial BMP image. The method and the device effectively combines the image conversion process from the JPG image to the BMP image with the object searching process carries out by utilizing the frequency domain information, greatly increases the searching and positioning speed of the image object and greatly reduces the calculated amount in the whole identifying process by partially converting the image object area into the BMP image and identifying on the BMP image, thereby obtaining the obviously effects of high seed and high efficiency.

Apparatus and method for 3D image conversion and a storage medium thereof

Abstract: An apparatus and method for three-dimensional (3D) image conversion and a storage medium thereof. The method implemented by a three-dimensional (3D) image conversion apparatus includes: receiving an input image including a plurality of frames; identifying a first frame selected among the plurality of frames; acquiring first depth information of a first object selected in the first frame; identifying a second frame selected among the plurality of frames; acquiring second depth information of a second object selected in the second frame; acquiring third depth information of a third object selected in a third frame, using at least one of depth information between the first depth information and the second depth information; generating fourth depth information based on the first depth information, the second depth information and the third depth information; and rendering the input image based on the fourth depth information.

2d to 3d image conversion

Abstract: A method (and system) of processing image data in which a depth map is processed to derive a modified depth map by analysing luminance and/or chrominance information in respect of the set of pixels of the image data. The depth map is modified using a function which correlates depth with pixel height in the pixellated image and which has a different correlation between depth and pixel height for different luminance and/or chrominance values.

Posterior sight support device for vehicle

Abstract: PROBLEM TO BE SOLVED: To obtain a sense of perspective in a display image showing a posterior part of a vehicle.SOLUTION: A door mirror camera 2 and a rearview mirror camera 3 capture posterior images (a right posterior image, a left posterior image, and a center posterior image) in the posterior part of its own vehicle A. Subsequently, an image conversion device 9 divides an area including a road surface in captured posterior images into multiple areas (a first area, a second area, and a third area) collaterally arranged from a far side of the vehicle A to a near side of the vehicle A. Then, the image conversion device 9 corrects posterior images to allow images of the areas on the farther side of the vehicle A to be unsharper among the multiple areas. A door mirror monitor 4 and a rearview mirror monitor 5 display corrected posterior images. Consequently, a sense of perspective is obtained in a display image showing the posterior part of a vehicle.

An approach to adjustment and reduction of the number of controlling parameters for simple 2D to 3D image conversion schemes

Abstract: In this paper an approach to adjustment and reduction of the number of parameters controlling simple 2D to 3D image conversion schemes is presented. With the reported experiments five 2D to 3D conversion methods were examined and compared. The controlling parameters were experimentally adjusted by the viewers. The results of experiments indicate a linear dependence between the parameters. Thus it is possible to reduce them to only one parameter.

Image conversion device, camera, video system, image conversion method and recording medium recording a program

Abstract: Provided are an image conversion device, a camera, a video system, an image conversion method and a program which are capable of performing a desired image conversion even when the orientation of multiple regions within one image is to be changed in different directions. This system is provided with: a region dividing unit (13) which divides one inputted image into multiple regions, and an image conversion unit (14) which converts the image of at least one of the regions created by the region dividing unit (13) to an image imaged from a virtual viewpoint different from the imaging viewpoint of the inputted image.

Object detection device

Abstract: The invention is provided with: an imaging means (10) that images a specified area; an image conversion means (31) that performs a point-of-view conversion on an image obtained by means of the imaging means, said image being converted to an aerial view image; a brightness detection means (40) that detects the brightness of the specified area; a first object detection means (33); and a second object detection means (37). Said first object detection means (33) generates difference waveform information from a difference image of aerial view images from different times, said aerial view images having been obtained from the image conversion means. If at least the brightness detected by the brightness detection means is at least a specified value, said first object detection means (33) detects an object by means of the difference waveform information being at least a first threshold value (α). Said second object detection means (37) detects edge information from an aerial view image obtained from the image conversion means. If at least the brightness detected by the brightness detection means is less than a specified value, said second object detection means (37) detects an object by means of the edge information being at least a second threshold value (β).

Image formation device and image formation method

Abstract: An image formation device of the present invention has: a recording medium supplying section that supplies a recording medium; a conveying section that conveys the recording medium supplied from the recording medium supplying section; an image formation section that ejects droplets and forms an image on the recording medium that is being conveyed; an image conversion section that converts an inputted image into dot data; a printing processing section that, from inputted print information, outputs continuously printed number of sheets information that causes the image formation section to continuously print output images, and prints the recording media corresponding to the continuously printed number of sheets, and, thereafter, carries out processing that temporarily stops printing of the output images; and a control section that, during a stoppage time of the printing, stops at least formation of images at the image formation section, and continues to drive the conveying section.

Parking assist device and parking assist method

Abstract: PROBLEM TO BE SOLVED: To provide a technique facilitating confirmation of an obstacle or the like around an own vehicle during parking.SOLUTION: A parking assist device comprises: an image conversion means for converting an image picked up by an image pickup means picking up an image around a vehicle into an over view image when viewing the vehicle side from a virtual viewpoint set above the vehicle and a back view image when viewing the vehicle side from a virtual viewpoint set behind the vehicle; and an image switching means switching to the back view image when a reverse travel signal indicating the selection of reverse travel as the traveling direction of the vehicle is detected and switching to the over view image when the reverse travel signal and a parking spot for parking the vehicle are detected.