Content-based representation for moth recognition and retrieval: a review

There are numerous of moths on the earth and their presence is useful to our life especially as life indicators. Most of the entomologists have problems to recognize moth’s species because each of them has their own color, texture and shape. The varieties of color, texture and shape of moths has increased the researcher’s attention to improve the method in recognizing the moth’s species. This study investigates the effectiveness of Bag of Visual Words (BOVW) representation for recognizing the moth’s species. BOVW is a simple and effective representation. This representation broadly used especially in computer vision and object recognition. Local descriptors, clustering approaches, and word-image histograms in regards to BOVW for image classification and retrieval are studied. There is a contradiction between BOVW models about spatial information. The extension of BOVW to consider the spatial information is believed able to contribute to a more effective representation for moth recognition and retrieval.

Text Content-based representation for moth recognition and retrieval.pdf Restricted to Repository staff only Download (123kB) Official URL or Download Paper: https://www.sciencepubco.com/index.php/ijet/articl...

Actions (login required)

View Item