Fast Automatic Compensation of Under/Over- Exposured Image Regions
This paper presents a new algorithm for spatially modulated tone mapping in Standard Dynamic Range (SDR) images. The method performs image enhancement by lightening the tones in the under-exposured regions while darkening the tones in the over-exposured,
- PDF / 2,115,516 Bytes
- 12 Pages / 430 x 660 pts Page_size
- 88 Downloads / 144 Views
Abstract. This paper presents a new algorithm for spatially modulated tone mapping in Standard Dynamic Range (SDR) images. The method performs image enhancement by lightening the tones in the under-exposured regions while darkening the tones in the over-exposured, without affecting the correctly exposured ones. The tone mapping function is inspired by the shunting characteristics of the center-surround cells of the Human Visual System (HVS). This function is modulated differently for every pixel, according to its surround. The surround is calculated using a new approach, based on the oriented cells of the HVS, which allows it to adapt its shape to the local contents of the image and, thus, minimize the halo effects. The method has low complexity and can render 1MPixel images in approximately 1 second when executed by a conventional PC. Keywords: Image Enhancement, Tone Mapping, Human Visual System.
1 Introduction Important differences often exist between the direct observation of a scene and the captured digital image. This comes as a direct result of the low dynamic range of the capturing device, compared to the dynamic range of natural scenes. Conventional SDR images (8-bits/channel) cannot acceptably reproduce High Dynamic Range (HDR) scenes, which is usual in outdoor conditions. As a result, recorded images suffer from loss in clarity of visual information within shadows (under-exposured regions), or near strong light sources (over-exposured regions). A straight-forward solution to this problem is the use of HDR capturing devices instead of the conventional SDR ones. Nevertheless, HDR cameras cannot always provide a practical solution. Their increased cost has limited their use, while the majority of the existing vision systems are already designed to use SDR cameras. Another possible solution is to acquire an HDR image by combining multiple SDR images with varying exposures [1]. However efficient, this solution is by its definition time consuming and thus, cannot be used in time-critical applications. Consequently, an unsupervised tone-enhancement algorithm for the under/over-exposured regions of SDR images could at least partially solve the problem, by enhancing the visual information of those regions, while minimally affecting the others. D. Mery and L. Rueda (Eds.): PSIVT 2007, LNCS 4872, pp. 510 – 521, 2007. © Springer-Verlag Berlin Heidelberg 2007
Fast Automatic Compensation of Under/Over-Exposured Image Regions
511
Many algorithms have been proposed in this direction the past decades. The most important of all is the Retinex family of algorithms. Retinex was first presented by Edwin Land in 1964 [2] and was motivated by some attributes of the HVS, which also defined its name (Retina & Cortex). The initial algorithm inspired several others and until today the most widespread version of Retinex is the Multi Scale Retinex with Color Restoration (MSRCR) [3], which has been extensively used by NASA and has been established in the market as commercial enhancement software (PhotoFlair). In MSRCR, the new p
Data Loading...
