Services on Demand
Article
Indicators
Related links
- Cited by Google
- Similars in Google
Share
South African Computer Journal
On-line version ISSN 2313-7835
Print version ISSN 1015-7999
Abstract
WINBERG, Simon; NAIDOO, Khagendra and RAMONE, Moeko. Accelerating computer-based recognition of fynbos leaves using a Graphics Processing Unit. SACJ [online]. 2017, vol.29, n.3, pp.238-262. ISSN 2313-7835. http://dx.doi.org/10.18489/sacj.v29i3.432.
The Cape Floristic Kingdom (CFK) is the most diverse floristic kingdom in the world and has been declared an international heritage site. However, it is under threat from wild fires and invasive species. Much of the work of managing this natural resource, such as removing alien vegetation or fighting wild fires, is done by volunteers and casual workers. The Fynbos Leaf Optical Recognition Application (FLORA) was developed to assist in the recognition of plants of the CFK. The first version of FLORA was developed as a rapid prototype in MATLAB, but suffered from slow performance and did not run as a lightweight standalone executable. FLORA was thus re-developed as a standalone C++ application and subsequently enhanced using a graphics processing unit (GPU). This paper presents all three versions, viz., the MATLAB prototype, the C++ non-accelerated version, and the C++ GPU-accelerated version. The accuracy of predictions remained consistent. The C++ version was noticeable faster than the original prototype, achieving an average speed-up of 42 for high-resolution images. The GPU-accelerated version was even faster achieving an average speed-up of 54. Such time saving would be perceptible for batch processing, such as rebuilding feature descriptors in the leaf database.
Keywords : computer vision; image processing; automated plant identification; parallel algorithms.