To see that there is an object in the image, you will want to determine where the edge of that object is. If you look at the image, you'll notice that there is a big difference in the color of the background compared to the color of the edge of each object.
You can have your program read in the color data, as you have done, and whenever it notices that the color has changed by a significant enough amount, the program will highlight this point. This would be done by applying a sobel filter to the color data.
Doing this should highlight the edges of the object. From here, you can try to analyze the size of the area enclosed or the features of these edges to determine the object that is highlighted.
Here is a link to a question asked about applying a sobel filter that has some good answers.
While libraries such as OpenCV do have sobel filtering and image detection, it does not take too much code to implement the sobel filter yourself. I believe it would be a more simple solution as well. Installing and learning OpenCV will likely take more time than just applying your own sobel filter.