BSc Honours Thesis

An Investigation into the Viability of Photo Identification System I3S as an Objective, Non-invasive tool to study Giraffa camelopardalis for Conservation Monitoring Strategies 


Full Thesis

You can download my entire thesis a a PDF here

Or download the chapters from the contents list.


Management of threatened species requires knowledge of the species population, social structures, behaviour and ranges, with data resulting from invasive and non-invasive techniques. With data gathered from four UK zoological collections, a study was conducted to test non-invasive photo identification software I3S; a program designed for use with the spotted, ragged tooth shark, Carcharias taurus for use with the patterned coat of the giraffe, Giraffa camelopardalis. Based on the program’s successes with ragged tooth shark, C. taurus, a success rate of 75% or higher, when using optimum images and 50% or higher when using sub-optimum images was expected, with a greater success rate when testing a gender specific database. Ten images were taken of 35 giraffe as a sample to test I3S as an accurate in-situ monitoring method for conservation, under optimum and sub-optimum conditions, which was judged by the angle of the animal within the image and light conditions. Each giraffe was tested four times, using one optimum and one sub-optimum image, testing the entire database, and the sex appropriated database, totalling 70 test results. When testing optimum images the success rate was 37%, and was the same for both databases. Sub-optimum images returned a success rate of 42% - 45% for both databases (chi-square 3.66453). This is significantly lower than was expected; however, statistical tests show that with a larger data set a higher accuracy rate would have been achieved. With more saved images of each profile, the program could determine more accurate results. Equally, the I3S software would benefit from working at more severe degrees such as in 3D, and by limiting the search to a particular side of the giraffe would illuminate a third of the database with each test, thereby increasing the accuracy rate.

Thanks to Jergen and Renate from I3S for letting me use their program.

I loved writing my thesis, and hope you enjoy reading it! My adviser Liz Webb was fantastic and it received 83% overall, which I was disappointed with at the time but a great learning curve and helped towards gaining a 1st class degree with honours.