This document summarizes a research paper that proposes a new method for segmenting text lines in machine printed Telugu script documents. It begins by noting common challenges with existing horizontal projection profile methods for line segmentation when text lines overlap. The proposed method first estimates the top and bottom bounding lines of the inter-line space using statistical analysis of the horizontal profile. It then predicts the segmentation path, which is set at 70% of the inter-line space to follow character contours. When two possible paths exist, it analyzes the horizontal profile of a small sub-image to select the correct path. The method achieved a success rate of 99.1% on tested documents.