#pragma GCC optimize("Ofast")
#include <bits/stdc++.h>
using namespace std;
#define ll long long
#define pb push_back
#define endl '\n'
#define AI(x) begin(x),end(x)
#ifdef DEBUG
#define debug(args...) LKJ("\033[1;32m[ "+string(#args)+" ]\033[0m", args)
template<class I> void LKJ(I&&x){ cerr << x << '\n'; }
template<class I, class...T> void LKJ(I&&x, T&&...t){ cerr << x << ", ", LKJ(t...); }
template<class I> void OI(I a, I b){ while(a < b) cerr << *a << " \n"[next(a) == b], ++a; }
#else
#define debug(...) 0
#define OI(...) 0
#endif
#define _ ios::sync_with_stdio(0);cin.tie(0);cout.tie(0);
signed main(){_
int n, m;
cin >> n >> m;
vector<vector<int>> g(n, vector<int>(m, 0));
vector<vector<int>> pre(n, vector<int>(m, 0));
int ans = 0;
for (int i = 0; i < n; ++i) {
for (int j = 0; j < m; ++j) {
char temp; cin >> temp;
g[i][j] = (temp == '#');
pre[i][j] = pre[i][j-1] + g[i][j];
}
}
for (int i = 0; i < n; ++i) {
for (int j = 0; j < m; ++j) {
if (!g[i][j]) continue;
int l = j - 1, r = j + 1, k = i + 1;
bool flag = 1;
while (l >= 0 && r < m && k < n) {
if (!g[k][l] || !g[k][r]) {
// cerr << "RRR";
// debug(i, j, l, r, k);
//debug(l, r, k);
break;
}
if (pre[k][r] - pre[k][l] > 1){
break;
}
l -= 1;
r += 1;
k += 1;
}
if (!flag) continue;
r -= 2;
l += 2;
while (k < n && r >= l) {
if (!g[k][l] || !g[k][r] || pre[k][r] - pre[k][l] > 1){
flag = 0;
// cerr << "KKK" << endl;
// debug(l, r, k, pre[k][r], pre[k][l]);
break;
}
r -= 1;
l += 1;
k += 1;
}
//debug(flag);
if (l <= r || k < i + 2) flag = 0;
ans += flag;
//if (flag) debug(i, j, l, r, k);
//cerr << "\n\n";
}
}
cout << ans << endl;
return 0;
}
