PavDoc - Human assisted iOS Application for analysis of pavement marking In this project, we present an iOS Application that attempts to minimize the subjectivity in assessing the color quality of pavement markings. It captures an image of a pavement marking and uses standard image processing techniques to objectively determine the percentage of the color or foreground to the pavement or background in a region of interest selected by the user. The system deploys adaptive image processing and connected components techniques to come up with a measure for the color quality in the selected image. Throughout the processing stages, the system recommends different thresholding values that the user can overwrite if they see fit. The system uses multiple languages and standard libraries including the OpenCV library in C++, Swift, and Objective-C. To test the system, we collected 200 yellow and white pavement marking images with a big variation in color quality. Some markings are freshly painted with crisp and clear colors and others have so fainted that makes the color very hard to detect. We ran many tests of the system using these images and we report our findings in this report.