Dense residual generative adversarial network capable of quickly removing rain

A network and residual technology, applied in biological neural network models, image enhancement, image analysis, etc., can solve problems such as excessive length and time-consuming residual rain lines, and achieve the goal of simplifying network structure, improving rain removal rate, and enhancing features. effect of contact
CN112258402APending Publication Date: 2021-01-22BEIJING INSTITUTE OF TECHNOLOGYGY
6 Cites 2 Cited by

Patent Information

Authority / Receiving Office
CN ยท China
Patent Type
Applications(China)
Current Assignee / Owner
BEIJING INSTITUTE OF TECHNOLOGYGY
Publication Date
2021-01-22

Smart Images

  • Figure 1
    Figure 1
  • Figure 2
    Figure 2
  • Figure 3
    Figure 3
Patent Text Reader

Abstract

The invention discloses a dense residual generative adversarial network capable of quickly removing rain. The dense residual generative adversarial network comprises the following steps: establishingan algorithm operation environment; b, establishing a rain removal data set; c, designing a dense residual generator sub-network for quickly removing rain; d, designing a dense residual discriminatorsub-network for quickly removing rain; e, designing a discriminator and a generator loss function; and f, performing image post-processing and index calculation, and verifying the rain removal effectof the algorithm model. According to the invention, the operation and test environment of the whole algorithm is built, and the time consumption is reduced as much as possible while the rain removal effect is ensured and enough background details are reserved through the designed rain removal model. According to a result test, the rain removing duration of each image is about 0.02 s, and the rainremoving efficiency is greatly improved.
Need to check novelty before this filing date? Find Prior Art

Description

technical field

[0001] The invention relates to the technical field of visual environment perception for drone driving, in particular to a dense residual generation confrontation network for fast rain removal. Background technique

[0002] For outdoor vision tasks, such as vision-based unmanned vehicle environment perception, pedestrian and road sign detection, tracking, road monitoring and other tasks, clean, clear and visible images are essential. Reliable visual images can help people complete various visual tasks more accurately and efficiently, while reducing or even avoiding unnecessary mistakes. However, under severe weather conditions, especially rainy weather, the visibility of images is seriously degraded, which brings great difficulties to various visual tasks. Therefore, the research on image deraining technology is imminent. However this task is very challenging. First of all, for heavy rain weather, rain will bring fog and blur the image, and it is difficult...

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