Contrast enhancement for color images using discrete cosine transform coefficient scaling
Arash Samani, Karen A. Panetta, Sos С. Agaian · 2016
For numerous security applications, the timeliness of having quality enhanced images are very critical. Therefore, it is essential to have a low-cost, real-time image enhancement and automatic evaluation tools that can be used be provided across many image based security platforms. In the past, there have been many algorithms proposed to improve image contrast, but most of these algorithms are implemented in the Spatial Domain and very few are designed to work in a Transform Domain. Furthermore, many existing contrast enhancement algorithms only improve the contrast of color images by enhancing the luminance of the image without improving the color content of the image. Besides, many of these methods are tailored to specific image settings. Furthermore, most current state of the art processors used in imaging systems today are equipped with built-in hardware Discrete Cosine Transform or Discrete Fourier Transform modules. In this article, a novel enhancement method to improve image contrast for grayscale and color images is introduced. The proposed method operates in the DCT domain and is computationally efficient, which makes it suitable for realtime imaging systems. Computer simulations demonstrate the efficacy of the proposed approach.