For the low visibility and contrast of the foggy image, I propose a single image defogging algorithm. Firstly, change the foggy image from the space of RGB to HSI and divide it into a plurality of blocks. Secondly, elect the maximum point of S component of each block and correct it, keeping H component constant and adjusting I component, so we can estimate fog component through bilinear interpolation. Most importantly, the algorithm deals with the sky region individually. Finally, let the RGB values of all pixels in the blocks minus the fog component and adjust the brightness, so we can obtain the defogging image. Compared with the other algorithms, its efficiency is improved greatly and the image clarity is enhanced. At the same time, the scene is not limited and the scope of application is wide.
Image enhancement is very important image preprocessing technology especially when the image is captured in the poor
imaging condition or dealing with the high bits image. The benefactor of image enhancement either may be a human
observer or a computer vision process performing some kind of higher-level image analysis, such as target detection or
scene understanding. One of the main objects of the image enhancement is getting a high dynamic range image and a
high contrast degree image for human perception or interpretation. So, it is very necessary to integrate either empirical or
statistical human vision psychology and perception knowledge into image enhancement. The human vision psychology
and perception claims that humans' perception and response to the intensity fluctuation δu of visual signals are weighted
by the background stimulus u, instead of being plainly uniform. There are three main laws: Weber's law, Weber-
Fechner's law and Stevens's Law that describe this phenomenon in the psychology and psychophysics. This paper will
integrate these three laws of the human vision psychology and perception into a very popular image enhancement
algorithm named Adaptive Plateau Equalization (APE). The experiments were done on the high bits star image captured
in night scene and the infrared-red image both the static image and the video stream. For the jitter problem in the video
stream, this algorithm reduces this problem using the difference between the current frame's plateau value and the
previous frame's plateau value to correct the current frame's plateau value. Considering the random noise impacts, the
pixel value mapping process is not only depending on the current pixel but the pixels in the window surround the current
pixel. The window size is usually 3×3. The process results of this improved algorithms is evaluated by the entropy
analysis and visual perception analysis. The experiments' result showed the improved APE algorithms improved the
quality of the image, the target and the surrounding assistant targets could be identified easily, and the noise was not
amplified much. For the low quality image, these improved algorithms augment the information entropy and improve the
image and the video stream aesthetic quality, while for the high quality image they will not debase the quality of the
image.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.