Bits counting algorithm (Brian Kernighan) in an integer time complexity

Can someone explains why Brian Kernighan's algorithm takes O(log N) to count set bits (1s) in an integer. A simple implementation of this algorithm is below (in JAVA)

int count_set_bits(int n){
    int count = 0;
    while(n != 0){
        n &= (n-1);
        count++;
    }
    return count;
}

I understand how it works by clearing the rightmost set bit one by one until it becomes 0, but I just don't know how we get O(log N).