Topic
Image conversion
About: Image conversion is a research topic. Over the lifetime, 2490 publications have been published within this topic receiving 19077 citations.
Papers published on a yearly basis
Papers
More filters
•
21 Jun 2013
TL;DR: In this article, the authors proposed a technique having small load of processing for precisely detecting sufficient number of feature points from an image having a wide non-focus area, where the information processing device for detecting feature points by analyzing a captured image comprises: detection means for detecting features from a first image after setting a first target value indicating number of features to be detected; conversion means for performing image conversion processing to the first image for acquiring a second image; and control means for setting a second target value based on number of the detected feature points, and the first target values, and making
Abstract: PROBLEM TO BE SOLVED: To provide a technique having small load of processing for precisely detecting sufficient number of feature points from an image having a wide non-focus area.SOLUTION: The information processing device for detecting feature points by analyzing a captured image comprises: detection means for detecting feature points from a first image after setting a first target value indicating number of feature points to be detected; conversion means for performing image conversion processing to the first image for acquiring a second image; and control means for setting a second target value based on number of the detected feature points, and the first target value, and making the detection means detect the feature points from the second image.
1 citations
•
31 Jan 2020
TL;DR: In this paper, an image semantic segmentation method based on discrete cosine transform (DCT) was proposed. But the method is not suitable for image decompression, and calculation consumption and time consumption can be avoided by removing a plurality of down-sampling operations.
Abstract: The invention discloses an image semantic segmentation method and device based on discrete cosine transform, and relates to the field of computer vision. The image semantic segmentation method based on discrete cosine transform comprises the following steps: converting an RGB image into DCT representation; rearranging the DCT coefficients by adopting FCR; and inputting the DCT representation dataafter coefficient rearrangement into an improved BiSeNet model to perform image semantic segmentation, the improved BiSeNet model being that a plurality of down-sampling operations are deleted on thebasis of the BiSeNet model, and the model depth is increased. The image semantic segmentation device based on discrete cosine transform comprises an image conversion module, a rearrangement module anda semantic segmentation module. According to the method, the RGB image of the original area is encoded into the component in the frequency domain through the DCT operation, and calculation consumption and time consumption caused by image decompression can be avoided.
1 citations
•
14 Feb 2013
TL;DR: In this article, an image conversion device and a method thereof are provided to generate an image as 3D content by selecting a plurality of images, where a matching unit extracts a matching point between the first and the second image and an image processing unit applies the first conversion parameter to the first image.
Abstract: PURPOSE: An image conversion device and a method thereof are provided to generate an image as 3D content by selecting a plurality of images. CONSTITUTION: When a first and a second image are selected, a matching unit extracts a matching point between the first and the second image(S910,S920). An image processing unit extracts a first and a second conversion parameter using the matching point(S930). The image processing unit applies the first conversion parameter to the first image. The image processing unit applies the second conversion parameter to the second image(S940). [Reference numerals] (AA) Start; (BB) End; (S910) Selecting an image; (S920) Extracting a matching point; (S930) Extracting first and second conversion parameters; (S940) Generating images on the left and right using the first and second conversion parameters
1 citations
05 Nov 2015
1 citations
•
27 Aug 2003
TL;DR: In this article, a radiographic image information reader includes an excitation and reading means to scan and erase the residual image of the radiographic conversion panel after completion of reading radiographic information.
Abstract: PROBLEM TO BE SOLVED: To provide a high image quality image by combining a CsBr stimulable phosphor, a semi conductor laser and a photomultiplier tube. SOLUTION: A radiographic image information reader includes an excitation and reading means to scan a radiographic image conversion panel having a stimulable phosphor layer with exciting light so as to photoelectrically read radiographic image information, and an erasing means to erase the residual image of the radiographic image conversion panel after completion of reading the radiographic image information. The stimulable phosphor layer consists of CsBr. The excitation and reading means and the erasing means are integrally reciprocated relative to the radiographic image conversion panel. The excitation and reading of the radiographic image information by excitation and reading means are carried out on the outward course in the reciprocative movement, and residual image erasing by the erasing means is carried out in the return course. COPYRIGHT: (C)2003,JPO
1 citations