Image conversion

About: Image conversion is a research topic. Over the lifetime, 2490 publications have been published within this topic receiving 19077 citations.

Image conversion method and device based on multi-cycle generative adversarial network

Abstract: The invention discloses an image conversion method based on a multi-cycle generative adversarial network, and the method comprises the steps: obtaining an image which needs to be converted, and dividing the image into a first image type and a second image type which are different; constructing at least one intermediate image category according to the images of the first image category and the images of the second image category; constructing a generative adversarial network comprising a plurality of generators and a plurality of discriminators according to the first image category, the secondimage category and the intermediate image category; wherein the first image category correspondingly generates a first discriminator, the second image category correspondingly generates a second discriminator, and the intermediate image category correspondingly generates a third discriminator; and performing image conversion through the generative adversarial network. According to the invention, the image conversion capability can be improved so as to complete more complex and difficult conversion tasks.

Image compositing device, image compositing method, and image compositing program

Abstract: PROBLEM TO BE SOLVED: To generate a proper panorama background image without influence of moving objects, even if an input image sequence contains not a scene where a moving object passes but a scene where the moving object is piled up as the last one. SOLUTION: An image conversion relation calculator 12 calculates a conversion relation between two images chosen from a photographed image sequence. Using the conversion relation, a pre-priority panoramic image generator 14 generates a pre-priority panoramic image, where priority is given to a pixel value of an image photographed before in time. A post-priority panoramic image generator 16 generates a post-priority panoramic image, where priority is given to a pixel value of an image photographed behind in time, using the conversion relation. When the pre-priority panoramic image is different from the post-priority panoramic image, a panoramic image candidate integrator 18 adopts the pixel area of the post-priority panoramic image for a part before in time and the pixel area of the pre-priority panoramic image for a part behind in time, as a panorama background image. COPYRIGHT: (C)2011,JPO&INPIT

Radiation image conversion panel and reading method of radiation image information

Abstract: PROBLEM TO BE SOLVED: To provide a radiation image conversion panel and a reading method of radiation image information capable of providing a high precision and high quality radiation image, regarding to a record and reproduction method of radiation image. SOLUTION: The radiation image conversion panel consists of a stimulable phosphor layer formed on a rigid sheet. When the radiation image conversion panel is supported by two edges facing each other or by finit points on the two edges and kept horizontally, the defection of the radiation image conversion panel is equal to or less than 1,000 μm. The reading method of radiation image information is provided by using the radiation image conversion panel. COPYRIGHT: (C)2003,JPO

Display image generation device, display image generation method, and display image generation program

Abstract: PROBLEM TO BE SOLVED: To reduce block noise through ordinary graphic processing even without providing any particular device. SOLUTION: A display controller for generating a display image containing a compressed image which is compressed and stored in the unit of a block beforehand in at least a part thereof, is provided. The display controller includes a compressed image acquisition part 134, a developed image conversion part 135, a filter image acquisition part 136, and a display image generation part 137. The compressed image acquisition part 134 acquires a compressed image. The developed image conversion part 135 converts the acquired compressed image into a developed image constituted of pixel values. The filter image acquisition part 136 acquires a filter image constituted of pixel values correspondingly to the compressed image. The display image generation part 137 calculates and combines the pixel values of the developed image and the pixel values of the filter image to generate a display image. Furthermore, the filter image comprises a noise component in an area corresponding to an area including the block boundary of the compressed image. COPYRIGHT: (C)2007,JPO&INPIT

Device and method for displaying bird's eye view image of around vehicle

Abstract: PROBLEM TO BE SOLVED: To provide a device for displaying a bird's eye view image around vehicle that enables a driver to recognize a solid obstacle near a joint, in a relatively simple manner and without imposing a processing load. SOLUTION: The device for displaying bird's eye view image around vehicle includes a plurality of on-vehicle cameras for taking images of the vehicle's perimeter; an image conversion processing means including lens distortion correction processing, visual point conversion processing, display region cutting-out processing, and a synthesizing means for connecting and composing the cut out display region to create a bird's eye view image around vehicle which are applied to images photographed by the cameras; and a display means for presenting the driver and vehicle occupants with the bird's eye view image. The image conversion processing means, further, includes a reference height change means for sequentially creating the bird's eye view images, with respect to a second to (n +1)th reference heights (n is an integer of 1 or larger) which are progressively higher than a road surface level, in addition to the road surface level (first reference height), and set a first to (n+1)th bird's eye view images. COPYRIGHT: (C)2010,JPO&INPIT