What is rows 2-18? Under the assumption that we can ignore them, just use textscan and take them as headerlines:
>> fid=fopen('sample.txt');
>> data=textscan(fid,'%f%f','headerlines',18);
>> fclose(fid);
>> width=data{1};
>> height=data{2};
However, only 100 rows are found then instead of 112. But as some of the rows before row 19 have more/less than 2 columns, they don't seem to have the same meaning.
