You are given a 2D grid of size $m \times n$, where each cell contains either a $0$ or a $1$.

A $1$ represents the home of a friend, and a $0$ represents an empty space.

Your task is to return the minimum total travel distance to a meeting point. The total travel distance is the sum of the Manhattan distances between each friend’s home and the meeting point.

The Manhattan Distance between two points (x1, y1) and (x2, y2) is calculated as:
|x2 - x1| + |y2 - y1|.

Constraints:

• $m ==$ grid.length

• $n==$ grid[i].length

• $1 \leq m, n \leq 50$

• grid[i][j] is either 0 or 1.

• There will be at least two friends in the grid.

The main idea of this algorithm is that the total Manhattan distance is minimized when all friends meet at the median position, calculated separately for rows and columns. As Manhattan distance can be split into vertical and horizontal components, we collect all the row indices and column indices of the friends and compute the distance to their respective medians. As we loop through the grid row-wise and column-wise, the row and column indices are gathered in sorted order naturally, so no additional sorting is needed. Finally, a two-pointer approach is used to efficiently compute the total distance by pairing positions from both ends toward the center.

Using the intuition above, we implement the algorithm as follows:

1. Create two vectors, rows and cols, to store all cells’ row and column indexes where grid[i][j] == 1.

2. Iterate through the grid row by row. For each cell that contains a 1, push the row index i into the rows vector.

3. Iterate through the grid column by column. For each cell that contains a 1, push the column index j into the cols vector.

4. Use the helper function getMinDistance(rows) to calculate the total vertical distance to the optimal row (median).

5. Use the helper function getMinDistance(cols) to calculate the total horizontal distance to the optimal column (median).

6. Return the sum of the two distances as the minimum total travel distance.

The getMinDistance helper function receives a list of positions, points, and returns the total minimum travel distance to the median. The points list contains either row or column indices of friends. As the Manhattan distance is minimized at the median, it uses a two-pointer technique as follows:

• Initialize a variable, distance, with $0$ to compute the total distance.

• Initialize two pointers, i and j, one at the start and the other at the end.

• Each step adds the difference points[j] - points[i] to the total distance.

• This process continues until the pointers meet.

• Returns the total computed distance.

Let’s look at the following illustration to get a better understanding of the solution: