A fair comparison among existing piece-wise linear approximation lossy compression schemes.

What is it about?

Comparisons of existing schemes can be misleading if no due care is paid to the exact cost, which we call the representation cost, which counts how many bits are required to represent the sequence of the produced line segments. We rectify the situation in the presented work by presenting a fair means to evaluate the representation cost of competing schemes. Schemes that produce a small number of line segments can result in a poor compression ratio when one considers the representation cost involved.

Featured Image

Photo by Mika Baumeister on Unsplash

Photo by Mika Baumeister on Unsplash

Why is it important?

The representation of line segments of existing piecewise linear approximation lossy compression schemes mostly relies on an often implicit expansion of the domain used to represent the sensed data and/or the domain of the time indices. on the floating-point domain. On the other hand, the sensor data is in a small integer domain that depends on the sensor's resolution. But there are a few schemes that rely entirely on the sensor data domain rather than on the assumption of the floating-point domain. It would be unfair unless we consider the exact representation cost of each line segment produced by the schemes for a fair comparison.

Perspectives