Patent Search Results

Adnan Khan and Noah Borel
11/30/15

Trends/Scopes: The field of using a compound eye has been deeply explored by various corporations. These groups tend to use similar models of a multi-aperture image capture apparatus but use different algorithms to differentiate themselves. Overall, the compound eye designs are with the purpose of tracking an object or multiple objects. In the field of object tracking patents, the trend is similar, as each group uses a different method for object tracking. Most claim that their own system is of great accuracy and effectiveness. 

Possible Areas of Innovation: Projects with the compound eye has not become greatly involved in real life applications. This comes into play with our own project, as we plan to involve the compound eye with our real life issue of tracking and taking down an unidentified drone. The next step is to learn from what is done with a working model and slowly bring it to reality. 

Weekly Progress Report #6

Weekly Progress Report #6
Adnan Khan and Noah Borel
11/29/15


Progress:
- Completed code for object tracking based on sequential frame difference.
   - Capture two consecutive images.
   - Convert captured RGB (color) images to gray scale (black to white) images.
   - Use a sensitivity value to change clarity of gray scale image.
   - Use openCV absolute difference function to create a single threshold image. Threshold image is a black and white image with moving object between the original frames represented with white pixels.
   - Use a blur value to expand/contract the threshold image's white pixel area.
   - Add crosshair to mark object. Display x and y position values of crosshair point.
   - Touch-Ups: Optimized the sensitivity and blur size to track the object easily.

- Completed patent search. Additional blog post will be made for finds on the search.

Problems:
- For the object tracking code, problems arose with the object bounding rectangle. The bounding rectangle was not able to be used, so a crosshair at the center of the object was used instead.

Plan:
- Post the patent search results.
- The next step for the object tracking is to learn how to find a moving object's velocity based on captured images.
- Revisit compound eye based on finds of patent search.

Weekly Progress Report #5

Weekly Progress Report #5
Adnan Khan and Noah Borel
11/22/15

Progress:
- Began work on code to object track based on sequential frame difference:
  -  Captures two images one after the other.
  -  Convert each frame from RGB (color image) to gray scale (black and white image) for absolute difference openCV function.
  -  Finds the absolute difference between the images: Purpose is to find what has changed between the frames. Threshold image is produced with only moving object frames represented by white pixels.

What I still have to do:
   -  Use blur function to make the threshold image clearer and more defined. Second threshold image produced.
   -  Middle of located object used for tracking. Bounding rectangle used around object's sides farthest from middle point.
   -  Display text and x and y position of object in frame.


Problems:
-  Parts of the code are confusing and hard to understand. It takes time to identify and research each line of code that is not understood. However, overall, the structure of the object tracking program is comprehended.

Plan:
-  Complete the object tracking with sequential frame difference program. This included the steps under "What I still have to do" in the progress section above.

Weekly Progress Report #4

Weekly Progress Report #4
Adnan Khan and Noah Borel
11/15/15


Progress:
-   Windows Visual Studio 2010 Professional was successfully downloaded to school computer (Lenovo Computer #10).
-   OpenCV 2.4.9 was successfully downloaded to same school computer.
-   OpenCV libraries were successfully linked to a demo Visual Studio project.
-   First project created to track object based on color (Explained in more detail in Progress Report #1): Steps:
    -   Original RGB image converted to HSV image.
    -   HSV image converted to binary image.
    -   Threshold values isolate targeted object to track based on HSV values.
    -   Erode and Dilate functions clarify the targeted object.
    -   OpenCV findContours function marks the targeted object and gives coordinate values.

Problems:
-   Learning how to download and set up the Windows Visual Studio took much more time than what was expected. Otherwise, everything else was completed smoothly.

Plan:
-   OpenCV: The color detection system was mainly just to make sure that the OpenCV was running correctly and that the platforms were set for future work. The next step is to object track without color by comparing two consecutive frames and searching for a difference between the images.
-   Compound Eye: Revisit the concept of the compound eye. Begin to brainstorm on ways to be able to process the curved image or multiple images.

Weekly Progress Report #3

Weekly Progress Report #3
Adnan Khan and Noah Borel
11/8/15

Progress:
- Tried to download Windows Visual Studio on computer from school.

Problems:
- There was trouble downloading Windows Visual Studio to the school computers. The versions that were downloaded were the most updated version and the one from 2010.
- The most updated version may not have worked as the computer was with an old OS.

Plan:
- Retry downloading Windows Visual Studio on computer from school.
- Download Windows Studio on home PC. Home PC is with a more updated OS so the modern Windows Visual Studio version may work on it.

Amcrest ProHD

Amcrest
- A highly rated security manufacturer.

Amcrest ProHD
- 1080p video at 30fps
- Relatively wide viewing angle of 90 degrees.
- 12 built in infrared LEDs for night vision use up to 32 feet
- Setup using iOS or Android device using Amcrest View Lite app
- Can be panned or tilted remotely using phone or computer
- Video can be streamed directly to computer, phone, tablet or NVR
- Built in motion alert system
- WiFi connection required for most features to work
- Requires power supply (no battery)

Weekly Progress Report #2

Weekly Progress Report #2
Adnan Khan and Noah Borel
11/1/15

Progress:
-  Research into OpenCV with Microsoft Visual Studio software or Python software.

OpenCV with Microsoft Visual Studio:
Installation and Creation Process:
-  https://www.youtube.com/watch?v=rXkrQeqZd8A
-  http://docs.opencv.org/2.4/doc/tutorials/introduction/windows_install/windows_install.html
-  https://www.youtube.com/watch?v=POpMQPM9YlY
-  https://www.youtube.com/watch?v=X6rPdRZzgjg
-  https://www.youtube.com/watch?v=bSeFrPrqZ2A

OpenCV with Python:
Installation and Creation Process:
-  https://www.youtube.com/playlist?list=PLEmljcs2yU0wHqeLlrytfuiqyNKTKIlOq
-  https://wiki.python.org/moin/BeginnersGuide
-  http://docs.opencv.org/3.0-beta/doc/py_tutorials/py_objdetect/py_face_detection/py_face_detection.html#face-detection


Problem:
-  Python seems to unfamiliar to use. It would need time to understand it better.
-  Visual Studio, even though having a complicated user interface, has more resources for instruction.

Plan:
-  Use Windows laptop from school and begin hands-on visual studio work.
-  Download both Visual Studio and OpenCV, and have the OpenCV libraries accessible to the software.
-  Test object tracking using color method.