Method for rapidly achieving image restoration processing based on GPU
A processing method and image technology, applied in the direction of image data processing, image enhancement, image analysis, etc., can solve the problems of difficult real-time calculation, slow running speed, and large amount of calculation in serial methods, and achieve rich details, improve speed, and improve performance. The effect of occupancy
Inactive Publication Date: 2013-11-27
CHINA CENT FOR RESOURCES SATELLITE DATA & APPL
View PDF4 Cites 11 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
The current MTFC processing research is based on the central processing unit (CPU). With the improvement of the spatial resolution and quantization resolution of remote sensing images, the amount of downloaded image data continues to increase. For satellite images with high spatial resolution MTFC Processing, due to the large amount of calculation, the running speed is slow and time-consuming, which has become a bottleneck in the real-time processing of ground system image restoration. Most of the remote sensing image processing algorithms are computationally intensive, and the serial method encounters difficulties in real-time calculation in application. Difficulties
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment
[0043] The inventive method can be realized in CUDA programming environment, as image 3 As shown in the CUDA programming mode, the CUDA parallel computing function running on the GPU is called a kernel. A complete CUDA program is composed of a series of device-side kernel function parallel steps and host-side serial processing steps, and the kernel is the entire CUDA A parallelizable step in a program. On the GPU side, the kernel function is organized in the form of a thread network (grid), and each grid is composed of several thread blocks (blocks), and each block contains many threads (threads), the number of threads in a kernel function It can reach thousands or even tens of thousands, so the number of threads running on the GPU at the same time is quite amazing. The initiation of GPU threads is lightweight, the system overhead of creating threads is very small, and the time spent on thread switching is also quite short.
[0044] Define the main kernel functions of the G...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
A method for rapidly achieving image restoration processing based on a GPU comprises the steps of (1) constructing a two-dimensional MTF matrix, (2) copying an image to be restored and the two-dimensional MTF matrix into a GPU cache, (3) loading the two-dimensional matrix into a sharing storage from the GPU cache, (4) segmenting the image to be restored into image blocks, (5) conducting rapid Fourier conversion on each image block on the GPU, (6) processing and dividing the GPU into a plurality threading blocks, conducting frequency domain filtering on one logical block in each threading block, (7) conducting Fourier inversion processing on each logical block after the frequency domain filtering, copying all the processed logical blocks out of the GPU cache, and completing the processing of the image restoration.
Description
technical field [0001] The invention relates to an image restoration processing method for remote sensing data with a large amount of data. Background technique [0002] When optical satellites acquire remote sensing images, they have to go through a series of links such as the atmosphere, optical system, and CCD. Each link may degrade the image, resulting in blurred images and a decline in image quality. In order to improve the definition and image quality of imaging, it is necessary to restore the data. The current MTFC processing research is based on the central processing unit (CPU). With the improvement of the spatial resolution and quantization resolution of remote sensing images, the amount of downloaded image data continues to increase. For satellite images with high spatial resolution MTFC Processing, due to the large amount of calculation, the running speed is slow and time-consuming, which has become a bottleneck in the real-time processing of ground system image...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More IPC IPC(8): G06T5/00G06T7/00
Inventor 朱晓波孙业超郝雪涛
Owner CHINA CENT FOR RESOURCES SATELLITE DATA & APPL