【Lintcode】631. Maximal Square II

题目地址：

https://www.lintcode.com/problem/maximal-square-ii/description

给定一个二维$0-1$矩阵$A$，求其所含的最大的左上到右下对角线（主对角线）全为$1$，其余位置全为$0$的方阵的面积。

思路是动态规划。我们考虑以$A[i][j]$为右下角的方阵，以它为右下角的满足条件的最大方阵的边长，取决于：
1、以$A[i-1][j-1]$满足条件的方阵边长；
2、其上方有多少个$0$；
3、其左方有多少个$0$。
设$f[i][j]$为以$A[i][j]$为右下角的方阵中满足条件的最大方阵的边长，设$g[i][j]$为$A[i][j]$上方的$0$的个数（包括$A[i][j]$自己），设$h[i][j]$为$A[i][j]$左方的$0$的个数（包括$A[i][j]$自己）。那么就有： g [ i ] [ j ] = { 0 , A [ i ] [ j ] = 1 1 + g [ i − 1 ] [ j ] , A [ i ] [ j ] = 0 h [ i ] [ j ] = { 0 , A [ i ] [ j ] = 1 1 + h [ i ] [ j − 1 ] , A [ i ] [ j ] = 0 f [ i ] [ j ] = { 0 , A [ i ] [ j ] = 0 1 + min ⁡ { g [ i − 1 ] [ j ] , h [ i ] [ j − 1 ] , f [ i − 1 ] [ j − 1 ] } , A [ i ] [ j ] = 1 g[i][j]=\begin{cases}0,A[i][j]=1\\ 1+g[i-1][j],A[i][j]=0\end{cases}\\h[i][j]=\begin{cases}0,A[i][j]=1\\ 1+h[i][j-1],A[i][j]=0\end{cases}\\f[i][j]=\begin{cases}0,A[i][j]=0\\ 1+\min\{g[i-1][j],h[i][j-1],f[i-1][j-1]\},A[i][j]=1\end{cases} g[i][j]={0,A[i][j]=11+g[i−1][j],A[i][j]=0​h[i][j]={0,A[i][j]=11+h[i][j−1],A[i][j]=0​f[i][j]={0,A[i][j]=01+min{g[i−1][j],h[i][j−1],f[i−1][j−1]},A[i][j]=1​代码如下：

public class Solution {
    /**
     * @param matrix: a matrix of 0 an 1
     * @return: an integer
     */
    public int maxSquare2(int[][] matrix) {
        // write your code here
        if (matrix == null || matrix.length == 0 || matrix[0].length == 0) {
            return 0;
        }
        
        int m = matrix.length, n = matrix[0].length;
        
        int res = 0;
        int[][] up0 = new int[m][n], left0 = new int[m][n], dp = new int[m][n];
        for (int i = 0; i < m; i++) {
            for (int j = 0; j < n; j++) {
                if (matrix[i][j] == 1) {
                    up0[i][j] = left0[i][j] = 0;
                } else {
                    up0[i][j] = left0[i][j] = 1;
                    if (i > 0) {
                        up0[i][j] += up0[i - 1][j];
                    }
                    if (j > 0) {
                        left0[i][j] += left0[i][j - 1];
                    }
                }
                
                if (matrix[i][j] == 0) {
                    dp[i][j] = 0;
                } else {
                    dp[i][j] = 1;
                    if (i > 0 && j > 0) {
                        int min0 = Math.min(left0[i][j - 1], up0[i - 1][j]);
                        dp[i][j] += Math.min(dp[i - 1][j - 1], min0);
                    }
                }
                
                res = Math.max(res, dp[i][j]);
            }
        }
        
    	// 求面积，要平方一下
        return res * res;
    }
}

时空复杂度$O(mn)$。