Alerting system for automatically detecting, categorizing, and locating animals using computer aided image comparisons

• US 9,750,238 B2
• Filed: 07/20/2015
• Issued: 09/05/2017
• Est. Priority Date: 07/23/2014
• Status: Active Grant

• Current Assignee: Metecs LLC
• Original Assignee: Metecs LLC
• Inventors: Huffman, Robert Kenneth, Jordan, Daniel David, Zalkin, Chadwick Jason
• Primary Examiner(s): Harandi, Siamak
• Application Number:
  US14/804,238
• Publication Number:
  US 20160026895A1
• Time in Patent Office: 778 Days
• CPC Class Codes: A01K 29/005 Monitoring or measuring activity, e.g. detecting heat or mating
• CPC Class Codes: A01M 31/002 Detecting animals in a given area
• CPC Class Codes: G06V 20/52 Surveillance or monitoring of activities, e.g. for recognising suspicious objects recognising microscopic objects G06V20/69
• CPC Class Codes: H04N 7/181 for receiving images from a plurality of remote sources
• CPC Class Codes: H04N 7/188 Capturing isolated or intermittent images triggered by the occurrence of a predetermined event, e.g. an object reaching a predetermined position signal generation from motion picture films H04N5/253

Abstract

Described are systems for (a) monitoring real-time animal activity in an area of interest using computer aided image comparison of a real-time image of the area and reference images of animals of interest and (b) providing informed alerts to a user.

Claims

1. A system comprising:

a camera mounted in a remote area;

a microprocessor connected to the camera;

a wireless communication unit connected to the microprocessor;

wherein the camera is configured to capture images that form of a photo stream or videos of an area of interest and load the photo stream or video those images to the microprocessor, where the camera captures a first photo stream or video of the area of interest and loads said first photo stream or video to the microprocessor, where the first photo stream or video features a plurality of frames, and where the camera captures a second photo stream or video and loads the second photo stream or video to the microprocessor;

wherein a computer code executes a high level logic flowchart to obtain a static state background or appearance of the area of interest as follows;

an image of the static state background or appearance of the area of interest is learned, wherein learning the static state background or appearance of the area of interest is accomplished via

(a) identifying the first video or photo stream in the microprocessor and (i) splitting each frame from the plurality of frames of the first photo stream or video into a grid of sections, (ii) comparing every section of a grid of a first frame with corresponding sections of the grids from every other frame of the plurality of frames to identify a minor variance in each grid section, (iii) storing each grid section with the associated minor variance, and (iv) stitching a static background together with the stored grid sections with associated minor variance, so that the static state background or appearance is defined by the stitched together grid sections with associated minor variance;

wherein the microprocessor is pre-loaded with reference images of animals of interest;

wherein the microprocessor features a computer code in the form of computer image comparison algorithms for comparing said second photo stream or video of the area of interest with the reference images of animals of interest;

wherein the second photo stream or video of the area of interest that is loaded on the microprocessor from the camera is compared with the static state background or appearance so that overlapping aspects of the static state background or appearance and the second photo stream or video are subtracted from each other to identify incongruities between the second photo stream or video and the static state background or appearance;

wherein at least one incongruity is identified and wherein the at least one incongruity is compared with the reference images of animals to determine whether the incongruity matches the reference images;

where a match is found between the incongruity and a reference image of an animal and the animal in the at least one incongruity is designated as the animal of the type corresponding to the animal of the reference image and categorized;

wherein the wireless communication unit is configured to send alerts; and

where a wireless signal is sent by the wireless communication unit to alert that an animal of interest has been spotted at the area of interest.

2. The system of claim 1 wherein the microprocessor is a smartphone.

3. A system for (a) monitoring animal activity in an area of interest using computer aided image comparison of at least one photo stream of images of the area of interest and a reference image of an animal of interest and (b) providing informed alerts to a user, where the system comprises:

a camera;

a computer operated microprocessor;

a wireless communication device;

a receiver;

wherein the camera, computer operated microprocessor, and wireless communication device are mounted in the area of interest that is remote relative to the receiver;

wherein the camera is connected wirelessly to the computer operated microprocessor;

where the camera is configured to provide both a first photo stream of images of the area of interest and a second photo stream of images of the area of interest to the computer operated microprocessor;

wherein the wireless communication device is configured to send alerts to the receiver;

wherein the microprocessor is loaded with the reference image of the animal;

wherein the computer operated microprocessor features a program algorithm for learning a static background image of the area of interest from the first photo stream of images of the area of interest, where the static background image represents an appearance of the area of interest without any moving objects;

wherein the program algorithm executes a high level logic flowchart to obtain the static background image of the area of interest as follows;

identifying the first photo stream of images in the computer operated microprocessor and (i) splitting each image of the first photo stream into a grid of sections, (ii) comparing every section of a grid of a first image of the first photo stream with corresponding sections of the grids from every other image of the first photo stream to identify a minor variance in each grid section, (iii) storing each grid section with the associated minor variance in the microprocessor, and (iv) stitching the stored grid sections with associated minor variance together so that the static background image is defined by the stitched together grid sections;

where the program algorithm executes on the computer operated microprocessor a comparison of the images of the second photo stream to the static background image to identify an incongruity between the images of the second photo stream and the static background image, where the images of the second photo stream of the area of interest are overlapped with the static state background or appearance so that overlapping aspects of the static background image and the images of the second photo stream are subtracted from each other to identify the incongruity;

where the program algorithm executes on the computer operated microprocessor a comparison of the incongruity and the reference images of the animal of interest and determines that the incongruity in a real time image represents the animal of interest;

where the comparison of the incongruity and reference image of the animal involve comparing shape, size, and color of the incongruity with shape, size, and color of the animal in the reference image of the animal;

wherein a wireless signal alerts the receiver of the determination that the incongruity in the real time image represents the animal of interest.

4. The system of claim 3 where the computer operated microprocessor is a smartphone.