Object recognition is the task of identifying distinct object in an image and typically associating the instance with a class of known objects. In this assignment you begin to experiment with a basic example of an object recognition task. Specifically, you will write and algortihm to find colored eggs in images. Your program will, by finding the individual eggs, be able to report how many are present in an image.

A series of images of colored eggs in four distinct contexts have been collected. Each image contains between 0 and 12 eggs. Here is a sample image shown at half normal resolution:

A total of 50 images have been collected, of which, 26 are being made available to you. The remaining images are being held back and will be used when grading your projects/programs. All 26 images may be viewed online.

A gzipped tar file of the images (42MB) is available for download at: eggs.tar.gz. The images are in three formats, JPEG and two versions of Portable Pixmap Format (PPM), use whichever makes the project easiest.

All 50 images are now available (5/4/2012 Ross Beveridge)

Your program will read a single file specifying a set of image file names, one file name per line. The file for these training images is:

• eggBasket.txt

Your program, which should have the name <your-last-name>CountEggs, will proceed to read each image file, and then write a line stating:

<image-name> contains <number> eggs

Realize that this interface is being defined to make the task of running your code on other images easy.

Please submit a single zip or tar file through RamCT containing everything necessary to compile and run your program on a department unix workstation.

If you have special reasons for not using, or at least porting your code, to the department workstations please work with in advance to come up with alternate arrangements.

Ross Beveridge 4/3/12