Your ‘curve’ has noise between x=0 and x=200 or so. You will likely need to exclude that region.
One option is to use interp1, for example:
xq = interp1(curve(idxrng), x(idxrng), lineval);
where ‘idxrng’ are the index values that exclude the noise, and ‘lineval’ is whatever y-value the line has.
Another option is to use polyfit to approximate the exponentially-descending part of the curve:
p = polyfit(x(idxrng), y(idxrng), 3);
xq = roots(p - lineval);
Without your data, it is not possible to write specific code.
