Easy access to video imaging, digital frame grabbing and image processing technology has opened up exciting opportunities in fluid mechanics education. This paper presents several examples of student work, and discusses new curricula based on these new developments. Complex, unsteady flows can now be quantified using multi-dimensional techniques. Image sequences can be stored and used in problem-solving. The author extrapolates these experiences to the hypothesis that image-based curricula can revolutionize fluid dynamics education, and permit the undergraduate to become familiar with fluids engineering technology, exploiting unsteady flows. INTRODUCTION In the hectic environment of a modern engineering school, it is difficult to conduct laboratory experiments suitable for hands-on participation by large classes. Advances in computational capabilities have created the demand for (and the relatively easy solution of providing) increased attention to computational fluid dynamics. At the same time, the time available to teach fluid dynamics is compressed by demands for "broader" curricula, advances in other fields, and the pressure to reduce the total hours required for the first degree. The laboratory course is an easy candidate for cuts, as Shop classes were in the past. There are two undesirable results: 1) The typical undergraduate gets very little exposure to the beauty, variety, and realities of fluid dynamics, instead seeing a collection of smart techniques for solving differential equations "governing" steady, laminar, attached, single-component flows of perfect gases over streamlined shapes. Sometimes this occurs at the expense of physical insight, and of students who might otherwise become intuitive leaders in fluid dynamics. 2) Prevalent opinion holds that the future is all in CFD, and that hardware experiments, if done at all, are purely for "code validation", with attention restricted to those phenomena which are already known and modeled in the codes.