Image snapping

Publisher Website
Google Scholar
Abstract
Cursor snapping is a standard method for providing precise point- ing in direct manipulation graphical interfaces. In this paper, we introduce image snapping , a variant of cursor snapping that works in image-based programs such as paint systems. Image snapping moves the cursor location to nearby features in the image, such as edges. It is implemented by using gradient descent on blurred ver- sions of feature maps made from the images. Interaction techniques using cursor snapping for image segmentation and curve tracing are presented This paper presents the methods for realizing image snapping. After an introduction to cursor snapping, we present techniques for searching an image for features. The task of identifying features is discussed with particular attention to edge detection. We then examine how cursor snapping applies to higher level tasks in image-based applications. 2. PRECISE POINTING IN DIRECT MANIPULATION Almost all object-oriented drawing, modelling and CAD programs provide some technique to aid the user with precise positioning. Bier (3) surveys approaches to this problem. Techniques aim to give the user the required precision, yet maintain the dynamic and free feel of direct manipulation. The most successful methods enhance direct manipulation positioning by guiding the position of the cursor from which drawing operations are carried out. In such schemes, the position of the software cursor is decoupled from the the screen location specified by the hardware pointing device, which we call the pointer. The cursor follows the pointer, but snaps to nearby locations of interest. Most snapping interfaces, with the exception of Venolia's 3D system (33), move the cursor discontin- uously: when the cursor gets within range, it jumps to its target. A common cursor positioning aid is the fixed grid. While easy to implement and use, it permits only operations that are grid-aligned. Grids could be used in an image editing program; however, to obtain precision with respect to features in an initial image, the fea- tures would have to be aligned with the grid. This is especially impractical when the images are acquired from the real world.

This publication has 0 references indexed in Scilit: