Use of root level logging interferes with other logging when imported

In essence, if we use the python logging facility, you should be careful because when you create a logging handler for a function, it seems that it is essentially forked off the root handler. Unless you remove the root handlers they remain active. This was a problem in my case because I imported the `ImageFileCollection` function and the moment I ran it for the first time, logging was activated including the root handler. This meant when *I* tried to use logging in my program, I was getting extra copies of the logging because the root handler was also responding to my logging calls.

I removed the root handlers by issuing the following:
```
# Turn off root loggers that are activated by ImageFileCollection
for handler in logging.root.handlers[:]:
logging.root.removeHandler(handler)
```

I would suggest unless you are intending to use the root handlers in `ccdproc` that you remove them yourself when you spin up the logging. That said, if there is another suggested way to handle this situation, I would appreciate the feedback.

Locate the ImageFileCollection entry point and inspect how it configures Python logging, especially the root handlers. Reproduce the import-and-run scenario, then verify that using ImageFileCollection does not leave root logging active or produce duplicate messages in the importing application's logs.

Written by the indexing model from the issue text.