Are you only getting 1 pulse per wheel turn? That can be a lot of lag between change in speed and data capture.
Could you try rolling a thin small roller that rolls on the bike wheel and watch -that- turn? Then you know of change within the circumference of the little wheel of when the change began.
If you can put a bar magnet across a wheel shaft end center and mount a Hall sensor right over that you can use analog read to know the angle of the wheel maybe 5 times a millisecond.
Do you know how many meters or feet 1 turn of your motorcycle wheel covers? Or how much the speed of that wheel can change in that distance?
There are problems you're going to have you don't even know yet. You really used delay(150)?
Take all your "good pictures" now so you can look back on those days later.