TLE in UCV2013H

[vaibhav_sof]

I am getting TLE in spoj.com/problems/UCV2013H/

I am trying to use Breadth first search type of algorithm and storing the results for the size of slicks in a map,however, it is quite slow. My code is as follows:

Got AC :D

Going by the AC submissions for the problem i think that may be there is some other algorithm. Can someone point out whether i need to improve this algorithm or my approach to the problem is wrong. Any help will be appreciated.
Thanks.

[leppyr64]

Your BFS is a little inefficient. You often push items into the queue more than once. You count some slicks more than once.

[vaibhav_sof]

Thanks for the quick reply leppy.

You often push items into the queue more than once.

In my algorithm, i check whether the adjacent element in a row or column contain a one(i check for a 1 to the left,right and directly below the current element), if they contain then only i push them into the queue. Also, afterwards i change the element's value to 2. So how can i possibly push the same item into the queue more than once?

You count some slicks more than once.

I didn't get this part. Could you please explain.

[leppyr64]

So how can i possibly push the same item into the queue more than once?

1 1
1 1
2,2 will get enqueued twice, once from 2,1 and once from 1,2.

I didn't get this part. Could you please explain.

You need to think harder about this part. Hint: all of the times you are counting the slick you are not getting the correct size.

[vaibhav_sof]

Thank you Leppy
You are awesome!