Rain Forest Canopy

A tropical rainforest is typically divided into four main layers: the emergent, canopy, understory, and forest floor layers. Of the four, it is quite tricky to distinguish the emergent layer from the canopy layer. The emergent layer contains a small number of very large trees, which grow above the general canopy, reaching heights of $45$ to $55$ metres. The canopy layer is the richest layer of the diverse rainforest, and ranges in heights of $30$ to $45$ metres tall. Because of biodiversity crisis, monitoring rainforest canopy is the mission of Whole Forest Observatory, which is trying to identify suitable canopy research sites.

A group of engineers have found a simple satellite imagery technique that can mark canopy layers on a spot image of a wide rainforest area. The spot image is a square and is stored as pixels, i.e., small cells containing either a 1 or a 0. Each pixel carries some information about a particular $1$ km^2 region. A pixel location contains a 1 if part or its entire represented region is a canopy layer and a 0 if otherwise.

The following assumptions hold:

• A canopy layer is represented by at least a single 1.

• Cells with adjacent sides on a common pixel that contains a 1, comprise the canopy layer. A single large canopy layer image will contain all 1's.

• Distinct canopy layers do not touch one another.

• There is no wrap-around, i.e., the pixels on the bottom are not adjacent to the top, and the left is not adjacent to the right.

Write a program that reads spot images and correctly counts the number of canopy layers in these images.