An algorithm for determination of the morphological characteristics of honey bees

Abstract

At the current level of science and technology is used semi-automatic measurement of body parts of the bees, yielding images taken with a reference object via a camera or a scanner and then perform measurement by software using a pointing device. There are attempts to fully automated process of measuring the morphological characteristics of bees, at this stage there are conversions for Measuring wings, but this process for other parts are still made by manual way. The informative colour features for the separation of tergite and probotics from background in the image are selected by distance functions and correspondence analysis. Distances are determined between the values of the colour components of the object and background. From statistical analysis is found that appropriate for the separation of an object from background are S and V colour components of the HSV colour model. Algorithms and program in Matlab environment for separating tergite and proboscis from the background of the image and definition of their main sizes are developed. From the analysis of the results is found that the major influence on the accuracy of the measurement is the angle at the disposal of the bee body part in the image.

Automatic Measurement of the Morphological Characteristics Of Honeybees With A Computational Program

“…Suitable devices for obtaining images are video camera, camera, and scanner (Lazarov, 2016). Regardless of popular literature attempts to fully automate the process of measuring the morphological characteristics of bees, at this stage, it is still done by manual way (Mattu et al, 1984;Roth et al, 1999;Tofilski, 2004;Mostajeran et al, 2006;Abou-Shaara et al, 2013;Zlatev et al, 2017).…”

“…The resulting distances are processed with the method of correspondences analysis (CA) (Kazlacheva, 2011;Kazlacheva et al, 2014) of the software Statistica. Informative colour features are determined on the basis of a certain available methodology (Georgieva et al, 2015;Mladenov et al, 2015;Dimitrova, 2016;Zlatev et al, 2017). The analysis of the developed algorithm aims to determine the extent to which its output variables are affected by moderate changes to the input data.…”

“…There are a number of methods for analyzing algorithms (Tofilski, 2011;Klingenberg, 2011;Abou-Shaara et al, 2013;Zlatev et al, 2017). One of these is by changing the input parameters by ± 10%.…”

“…These deviations, squared and appropriately scaled, are summed up to yield the chi-squared statistic in C. The data processing steps with the Correspondence Analysis method are presented in Table 3. The influence of the angle of rotation of the object on the measurement accuracy is checked by means of the relative measurement error module ԑ,%, which is determined by the following relationship (Klingenberg, 2011;Georgiev et al, 2014;Zlatev et al, 2017):…”