We focus on ten societal areas.
TNO cooperates with companies, the public sector and other organisations, to apply our knowledge and expertise with and for others.
We develop knowledge not for its own sake, but for practical application.
TNO offers you the chance to do groundbreaking work and help customers and society with innovative, practical and smart solutions.
On TNO Insights you can read in-depth interviews and articles. Get inspired by TNO’s innovations!
Existing techniques for increasing visual contrast suffer when single images contain various types of borders. The existence of strong edges may cause weak edges to become lost in the contrast enhancement process. In particular, the contrast function can be guilty of wiping out small artifacts near large borders. Changing the threshold for contrast may cause an excess of noise in other areas. The invention proposes a technique for increasing the visibility of both weak and strong edges.
The invention uses a localizing algorithm to improve contrast on visual displays. It compares each pixel to its neighbors and then assigns weight to each on the basis of local similarity. Pixel weights are used to determine the edge locations, which are then used in a modification of the whole image's contrast. This system will increase the visibility of weak edges, and can be supplemented with the secondary use of other filters that affect strong edges equally. This contrast function can be performed, as is typical, for intensity, but also for a specified color axis.
The algorithm analyzes small-scale statistical data in order to determine whether an edge is present. It also seeks to determine whether a gradient exists and, if not, to increase the strength of the contrast effect. This calculation is made on the basis of the similarity—for whatever value—of each pixel to its neighbors. The results of this calculation inform the image editor to transform each pixel within a specified range. The user may define a 'stretching factor,' which specifies the relationship between the inputted difference between two pixels and the resulting output image.