Freehand Drawing Segmentation Based on Color Differences

Abstract

Abstract Image segmentation is a fundamental step in computer vision problems such as object detection, localization, recognition, etc. The accurate segmentation has a vital effect on the output results. n this work, we introduce a new approach of segmentation based on freehand drawing on the target object on the input image then segmented that object from the image based on the color differences in the image regions. The method applied to different types of images to justify the robustness of the method. The experimental results show the effectiveness of the proposed method to segment the freehand drawn object with different characteristics of the images (lightness, background, number of objects).