MSEE thesis. Financial support from the Indiana CST, Purdue Plant Biotechnology Center.


Concealed insect infestations in stored grain are responsible for large economic losses worldwide, and have traditionally been difficult to detect arid quantify. A simple method of infestation detection and quantification has been developed, taking advantage of the ultrasonic emissions generated by the feeding activity of the insects. The acoustic signals generated by the insects are characterized as bursts of energy in the frequency range of 5 to 75 kHz, varying with the variety of seed being examined. These signals can be detected reliably by monitoring the seeds for sounds in a frequency band between 30 and 50 kHz. The stage of development of an insect, and thus the amount of damage of which the insect is capable, can be predicted by studying a time series of these signals. A basic signal acquisition procedure has been developed which amplifies the variations in the pattern of acoustic signals associated with different stages of larval development for the cowpea weevil. A histogram is constructed to describe the time intervals between successive feeding events, and is compared to typical histograms associated with each stage of development. In over 80% of the cases, an acquired histogram from a cowpea weevil at a known stage of development was most similar to the typical histogram associated with insects at the same stage of development. Using this correlation to quantify an infestation could lead to a significant reduction in the use of pesticides for insect eradication.

Date of this Version