A signal processing method for extracting vibration signals due to ants' activities

Publication Type:
Conference Proceeding
42nd International Congress and Exposition on Noise Control Engineering 2013, INTER-NOISE 2013: Noise Control for Quality of Life, 2013, 5 pp. 3631 - 3640
Issue Date:
Filename Description Size
InterNoise2013_final.pdfPublished version1.84 MB
Adobe PDF
Full metadata record
Many software algorithms have been developed to track ants by analysing recorded videos. On the other hand, the feasibility of using vibrations measured at the substrate to classify ants' behaviour has not been examined before. A method is developed to separate vibrations owing to ants' activities from the substrate's response through a filtering/de-convolution procedure. This involves estimating the frequency response of the substrate and applying wavelet analysis to the measured vibrations. A number of responses due to ants' behaviours have been observed: Ants shaking, falling, carrying stones, walking, scratching/biting, tapping hind legs, grooming, and antennation/feeding. Vibrations produced by ants falling, carrying stones, walking and scratching/biting are measurable (i.e, above background noise levels). The proposed method is shown to be successful in classifying activities due to ants falling, ants carrying stones and to a lesser extent ants' scratching/biting. With further refinement, it seems feasible to use vibrations and the proposed algorithm to measure ants' behaviours in bioassays. Copyright© (2013) by Austrian Noise Abatement Association (OAL).
Please use this identifier to cite or link to this item: