How to normalize data along the x-axis in order to perform cross-correlation?

I am working with EMG data for gait analysis. I am attempting to determine where in the gait cycle (heelstrike to heelstrike) the peak EMG signals change timing. In other words, does the peak occur at the beginning or end of the gait cycle compared to a baseline walking trial.

My EMG signal is "cut" into segments based on the time duration of each stride. From there, I need to make each EMG signal the same length (100 frames would work best for use with percentages) in order to perform cross-correlation and determine difference in EMG timings.

However, I am having a hard time determining a way to make each "cut" EMG signal the same length. I would have to alter the signal along the time (or frame number) domain and not the amplitude, which I know how to do.

I have attempted to use the interp1() function to no avail, since when lengthening, there is no end value to use to interpolate. If anyone has done anything like this before or can think of a creative way to make sets of data with differing lengths equal while preserving signal please let me know.