Vast libraries of historic photographs are currently degrading due to the effects of semi-transparent water blotches. Most current restoration techniques involve heavy user interaction and are therefore too expensive to use for large quantities of images. This paper introduces the Localized Logarithmic Image Restoration Algorithm, which provides an automated system to restore the water blotches found in old photographs. The Algorithm utilizes a new localized image processing framework that allows it to improve upon existing restoration methods. This framework can be used for a variety of image processing applications. In the presented application of blotch removal, new logarithmic and statistical equations are introduced and used within the localized framework to complete the restoration. As shown by intensive computer simulations, the Algorithm produces enhanced results when compared to the existing automated algorithm. Improvements include superior edge removal around the blotch, better local contrast preservation, and expanded saturation reduction capabilities. In addition to the enhanced restoration quality of the results, the simulations are also promising with respect to computational time.