首页> 外国专利> IMAGE QUALITY-ENHANCEMENT LEARNING DEVICE AND IMAGE QUALITY-ENHANCEMENT DEVICE

IMAGE QUALITY-ENHANCEMENT LEARNING DEVICE AND IMAGE QUALITY-ENHANCEMENT DEVICE

机译:图像质量增强学习设备和图像质量增强设备

摘要

To provide an image quality-enhancement learning device and an image quality-enhancement device capable of efficiency achieving high image quality enhancement with a small processing load.SOLUTION: An image quality-enhancement learning device includes: region extraction means 11 which extracts an adjacent region (c) including a pixel A of interest and a peripheral region (b) including the pixel A of interest from a resolution conversion image (a); approximate pixel specifying means 12 which acquires a pixel on a high-image-quality image (d) corresponding to the pixel A of interest as a target pixel R, and specifies a pixel having a pixel value closest to that of the target pixel among pixels in the adjacent region as an approximate pixel; and learning mans 13 which has a neural network constitution having an input layer, a plurality of intermediate layers, and an output layer, wherein the input layer comprises means for capturing respective pixels included in the peripheral region, the output layer comprises neurons, the number of which corresponds to the respective pixels included in the adjacent region, and a coefficient between neurons between the layers is changed such that a value of a neuron corresponding to the approximate pixel of the output layer is changed.SELECTED DRAWING: Figure 2
机译:为了提供一种图像质量增强学习设备和图像质量增强设备,其能够以较小的处理负荷高效地实现高图像质量增强。解决方案:图像质量增强学习设备包括:区域提取装置11,其提取相邻区域(c)包括来自分辨率转换图像(a)的感兴趣的像素A和包括感兴趣的像素A的外围区域(b);近似像素指定装置12,其获取与关注像素A相对应的高图像质量图像(d)上的像素作为目标像素R,并在像素中指定像素值最接近目标像素的像素。在相邻区域中作为近似像素;学习人13,其具有神经网络结构,该神经网络结构具有输入层,多个中间层和输出层,其中,输入层包括用于捕获包括在外围区域中的各个像素的装置,输出层包括神经元,数量图2中的像素对应于相邻区域中包含的各个像素,并且层之间的神经元之间的系数发生更改,从而与输出层的近似像素相对应的神经元的值也发生了变化。图2

著录项

相似文献

  • 专利
  • 外文文献
  • 中文文献
获取专利

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号