Apparatuses And Methods For Bio-Sensing Using Unmanned Aerial Vehicles

Abstract

Described herein are methods, apparatuses, and systems that enable a light weight autonomous unmanned aerial vehicle (UAV) to process hyperspectral (HSI) data during its flight and send information to the ground computer via radio-link. This capability is not currently available owing to the severe combination of technical constraints: the typical processing power required to analyze HSI data in real time; the small space and power available for the payload in a light-weight UAV; and the limited bandwidth available on the wireless link.

What is claimed is: 
     
         1 . A method for bio-sensing comprising:
 within an unmanned aerial vehicle comprising an optionally-detachable first general purpose digital computer system and an optionally-detachable hyperspectral imager, making a first special-purpose digital computer system by storing a first executable application program in a memory of the first general purpose digital computer system, and executing the stored program to impart to the first general purpose computer system the functionality of at least analyzing data from the hyperspectral imager, by changing the state of a one or more processors within the first general purpose computer system when aerial vehicle program instructions are executed; and   flying the aerial vehicle, including the hyperspectral imager, over at least a portion of a field of crops which are arranged in the form of a plurality of plant rows, the plant rows separated in an alternating pattern by a plurality of soil regions, while executing the aerial vehicle program instructions;   obtaining a priori information comprising a spectral signature for said plant rows and a spectral signature for said soil regions;   obtaining from the hyperspectral imager a series of images of the portion of the field of crops, comprising an image of the plurality of rows, wherein a first plurality of the pixels are generated by light obtained solely from the plant rows, and a second plurality of the pixels are generated by light obtained solely from the soil regions, both first and second pixels comprising spectral information in at least 16 bands;   wherein the program instructions comprise:
 classifying a plurality of pixels of the series of images into at least the categories of plant row or soil region, based on applying a first classifier algorithm to the series of images, the first classifier algorithm having being trained by said a priori information; 
 using spectra of a plurality of pixels categorized as plant row, applying a second classifier algorithm to said spectra to classify the spectra according to a predetermined classification relating to the health or quality of the crops; and 
   further comprising transmitting results of the classification wirelessly from a transmitter in the aerial vehicle to a receiver on the ground.   
     
     
         2 . The method of  claim 1 , wherein the hyperspectral imager is detachable, further comprising:
 performing a calibration, using the hyperspectral imager, by moving the imager at a known height less than 10 meters over a plurality of plant rows, thus capturing a plurality of close-up images over time of the plant rows and the soil regions, wherein said spectral signature for said plant rows, and said spectral signature for said soil regions, are obtained from the close-up images; and   training said first classifier algorithm using said a priori information.   
     
     
         3 . The method of  claim 2 , wherein moving the imager comprises attaching the imager to a line or cable passing over the plant rows and the soil regions, and moving the imager from one side of the line or cable to another. 
     
     
         4 . The method of  claim 2 , wherein the step of performing the calibration is performed within 30 minutes of flying the aerial vehicle. 
     
     
         5 . The method of  claim 2 , wherein the spectral signature for said plant rows is a signature of a first part of a plant comprising the plant rows, and wherein the a priori information further comprises a spectral signature for a second part of the plant. 
     
     
         6 . The method of  claim 2 , wherein the first part of the plant comprises leaves, and the second part of the plant comprises a fruit or vegetable. 
     
     
         7 . The method of  claim 1 , wherein the receiver on the ground is in communication with a second general purpose digital computer system, which has been made into a second special-purpose digital computer system by storing a second executable application program in a memory of the second general purpose digital computer system, and executing the stored program to impart to the second general purpose computer system the functionality of at least controlling one or more actuators on the ground, by changing the state of a one or more processors within the first general purpose computer system when aerial vehicle program instructions are executed, wherein the state of the one or more actuators is determined by the results of the classification. 
     
     
         8 . The method of  claim 7 , wherein the one or more actuators comprise an irrigation valve. 
     
     
         9 . A bio-sensing system comprising:
 an unmanned aerial vehicle comprising an optionally-detachable hyperspectral imager;   a wireless transmitter;   an special-purpose digital computer system comprising:
 an optionally-detachable first general purpose digital computer system comprising one or more processors and a memory comprising one or more digital storage devices; 
 a first executable application program stored in the memory, configured to be executed by the first general purpose computer system, imparting to the general purpose computer system the functionality of at least analyzing data from the hyperspectral imager, by changing the state of the one or more processors when aerial vehicle program instructions are executed, wherein the aerial vehicle program instructions comprise:
 obtaining from the hyperspectral imager a series of images of a portion of the field of crops obtained by flying the aerial vehicle, including the hyperspectral imager, over at least a portion of a field of crops which are arranged in the form of a plurality of plant rows, the plant rows separated in an alternating pattern by a plurality of soil regions, while executing the aerial vehicle program instructions; wherein the series of images comprise an image of the plurality of rows, wherein a first plurality of the pixels are generated by light obtained solely from the plant rows, and a second plurality of the pixels are generated by light obtained solely from the soil regions, both first and second pixels comprising spectral information in at least 16 bands; 
 classifying a plurality of pixels of the series of images into at least the categories of plant row or soil region, based on applying a first classifier algorithm to the series of images, the first classifier algorithm having being trained by a priori information comprising a spectral signature for said plant rows and a spectral signature for said soil regions; 
 using spectra of a plurality of pixels categorized as plant row, applying a second classifier algorithm to said spectra to classify the spectra according to a predetermined classification relating to the health or quality of the crops; and 
 causing the unmanned aerial vehicle to transmit results of the classification wirelessly from the transmitter in the aerial vehicle to a receiver on the ground. 
 
   
     
     
         10 . The system of  claim 9 , further comprising a line or cable passing over the plant rows and the soil regions, wherein the aerial vehicle comprises a means for slidably engaging with the line or cable. 
     
     
         11 . The system of  claim 9 , further comprising:
 one or more actuators; and   
       a second general purpose digital computer system, which has been made into a second special-purpose digital computer system by storing a second executable application program in a memory of the second general purpose digital computer system, and executing the stored program to impart to the second general purpose computer system the functionality of at least controlling the one or more actuators on the ground, by changing the state of a one or more processors within the first general purpose computer system when aerial vehicle program instructions are executed, wherein the state of the one or more actuators is determined by the results of the classification. 
     
     
         12 . The system of  claim 9 , wherein the one or more actuators comprise an irrigation valve.