Scale invariant feature approach for insect monitoring

Optical identification of bumblebee species: Effect of morphology on wingbeat frequency

“…Different technologies have been investigated for the (semi-) automatic detection and classification of insects. A first approach uses image processing to classify insects (Weeks et al, 1999;Arbuckle et al, 2001;Solis-Sánchez et al, 2011). One approach involves manual photography of a part or the complete insect, which has to be fixed to get good images (Weeks et al, 1999;Arbuckle et al, 2001;Solis-Sánchez et al, 2011).…”

Automatic identification and counting of small size pests in greenhouse conditions with low computational cost

“…In order to deal with the high quality images, more sophisticated algorithms would be necessary such as scale invariant feature transformation (SIFT) (Solis-Sánchez et al, 2011). However, since the appearance of three species (thrips, aphids and whitefly) were mainly distinguishable with color features in field conditions, the complicated algorithms applied to high quality images were not necessary to detect and identify pests collected on sticky traps in field conditions in practical aspect as demonstrated in this study.…”

“…SolisSánchez et al (2011) utilized scale invariant feature transform (SIFT) as a feature descriptor of the segmented insect images, combining with shape feature of insect body. These methods, however, in general also require either high resolution of images (Cho et al, 2007;Solis-Sánchez et al, 2011) or computationally expensive recognition algorithms (e.g., Gabor filter, SIFT). Although high pest recognition accuracy was achieved by these methods, the methods are in practical aspect not specifically suitable for large scale, long term pest monitoring (e.g., agricultural sensing network) in field conditions due to high computational cost.…”

Artificial neural networks in online semiautomated pest discriminability: an applied case with 2 Thrips species

“…Undisputedly, the use of artificially intelligent systems has spread to many fields of science (Weeks and Gaston, 1997;Do et al, 1999), including applied entomology (Solis-Sanchez et al, 2001;Fedor et al, 2009). ANNs, using a highly interconnected group of simulated neurons that process information in parallel (Haralabous and Georgakarakos, 1996) and learning from a set of examples, have been widely theoretically described (Freeman and Skapura, 1992;Ripley, 1993;Haykin 1994;Haralabous and Georgakarakos, 1996); therefore, there is no need to do so in this methodical paper.…”