Granted Patent B2

US 12,610,935 · App. 17/496,563 · Granted Apr 28, 2026

Inventors: Sriram Chellappan (Tampa, FL); Stephen Edward Saddow (Land O Lakes, FL); Ryan Marc Carney (Temple Terrace, FL); Brandon Wolfram (Seffner, FL); Mark Weston (Tampa, FL)

Assignee: University of South Florida

A01M1/026A01M1/023G06N3/04G06N3/08G06V10/95G06V20/00H04R1/08

Abstract

Claims (34)

1 . An apparatus to catch an insect and gather identifying data regarding the insect, the apparatus comprising:

a lamb-skin membrane body positioned within a membrane enclosure configured to mimic human skin and support the insect thereon, wherein the lamb-skin membrane body is fillable with a carbon dioxide gas attractant and porous to allow the carbon dioxide gas attractant to slowly leak out, the lamb-skin membrane body connected to a brass puck comprising a cartridge heater for heating the brass puck, wherein the brass puck is positioned in proximity to an inlet opening for directing the insect to perch on the lamb-skin membrane body and seek the carbon dioxide gas attractant disposed within the lamb-skin membrane body;

a plurality of imaging devices having respective fields of view that encompass the lamb-skin membrane body;

a microphone having a user-adjustable frequency range for increased or decreased sensitivity positioned within the apparatus and configured to obtain audio signals;

an active infrared sensor positioned at the inlet opening, wherein data obtained from the active infrared sensor is used to trigger the microphone to obtain the audio signals;

at least one sensor on an external surface of the apparatus configured to capture ambient environmental data including at least one of temperature, air pressure, altitude, humidity, latitude, longitude, time, CO 2 concentration and ambient light conditions; and

a processor operatively coupled to the plurality of imaging devices, the processor being operably connected to computerized memory storing computer-executable instructions which when executed by the processor, cause the processor to:

receive the audio signals from the microphone;

calculate a frequency signal corresponding to a wingbeat frequency of the insect;

responsive to determining that the wingbeat frequency is within a predetermined frequency range, trigger, via the active infrared sensor, the plurality of imaging devices to obtain images of the insect from a plurality of different angles,

determine a size of the insect based on a wing-beat frequency determined from the audio signals,

determine at least one anatomical component of the insect from the obtained images;

classify, using a machine learning classifier, the insect into a genus and species based, at least in part, on the determined size of the insect and the at least one determined anatomical component,

store insect data and the ambient environmental data in the computerized memory, and

responsive to detecting that the insect cannot be identified, generate an alert and transfer at least a portion of the stored data including the insect data and the ambient environmental data to a digital repository,

wherein the data is used to generate a set of training images for training one or more machine learning algorithms to identify the insect.

2 . The apparatus of claim 1 , wherein the membrane enclosure surrounds the lamb-skin membrane body to trap the insect within the membrane enclosure.

3 . The apparatus of claim 2 , further comprising an enclosed chamber connected to the membrane enclosure with an exit opening and configured to house the insect after the data has been stored.

4 . The apparatus of claim 1 , wherein the brass puck is connected to a source of carbon dioxide, and wherein the carbon dioxide enters the lamb-skin membrane body through the inlet opening.

5 . The apparatus of claim 1 , wherein the brass puck defines a recess for placement of additional insect attractants.

6 . The apparatus of claim 1 , wherein the plurality of imaging devices comprises a passive infrared sensor (PIR).

7 . The apparatus of claim 1 , wherein the plurality of imaging devices comprises a camera.

8 . The apparatus of claim 1 , further comprising a plurality of enclosed chambers connected to the membrane enclosure via respective exit openings and configured to house mosquitoes in respective ones of the plurality of enclosed chambers according to a mosquito taxonomy.

9 . The apparatus of claim 1 , wherein the computerized memory has computer-executable instructions stored thereon that, when executed by the processor, cause the processor to classify the insect using at least one neural network algorithm applied to image data obtained from the plurality of imaging devices.

10 . The apparatus of claim 8 , further comprising a funnel connected to the membrane enclosure to direct the insect into the apparatus.

11 . The apparatus of claim 1 , wherein the computerized memory has computer-executable instructions stored thereon that, when executed by the processor, cause the processor to:

determine the at least one determined anatomical component using at least one of a mask region-based convolutional neural network and a pixel-wise segmentation operation.

12 . The apparatus of claim 1 , wherein the computerized memory has computer-executable instructions stored thereon that, when executed by the processor, cause the processor to further:

output, to an application executing on a mobile device that is in electronic communication with the processor, at least a portion of the stored data for display.

13 . The apparatus of claim 1 , further comprising metal plate within the apparatus operatively coupled to a Peltier module configured to freeze the insect.

14 . The apparatus of claim 1 , wherein the processor is configured to determine the at least one anatomical component of the insect from the obtained images by:

identifying body pixels from the image data;

converting the body pixels into a selected color space data set; and

identifying textural features from the selected color space data set.

Assignments (1)

Continuity (2)

References Cited (63)