It is impossible to calculate and here's why.
1) The basic mathematics involved: - Newtons Universal law of Gravitation.
F = GMm/R^2
where:
F is the force of attraction between two objects
G is the universal gravitational constant; G = 6.67*10-11 N-m2/kg2. The units of G can be stated as Newton meter-squared per kilogram-squared or Newton square meter per square kilogram.
M and m are the masses of the two objects
R is the distance between the objects, as measured from their centers
GMm/R^2 is G times M times m divided by R-squared
Why you can't calculate it from a video clip.
1) You have no direct way of measuring the
respective masses of the objects involved.
2) in respect of the video in question (not seen it but i'm sure I know which clip it is) the vehicle is not parallel to the camera, thus rendering it almost impossible to accurately calculate the respective distances between the rover, dust and ground plane particularly at oblique angles relative to the camera.
3) In a normal video the timeframe simply isn't accurate enough for accurate time measurement better than approx 1/30th of a second. Thus inducing enormous errors in to the calculation. (this is based on a 29-30fps rate).
4) The field of view of the camera that took the video is unknown, nor from a video clip alone can you deduce any accurate measurement unless there is an accurate measurement scale or 3D grid imposed on the clip at the time of recording. Not knowing the field of makes all distance measurement at best a rough approximation and at worst an outright guess.
Addendum.
In response to the two posters below me:
Yes you can calculate it if you bring in external information that is NOT included in the video clip.
However, the posters original question is actually "Is it possible to extract the value of g (the gravitational constant of a planet) from a video clip?"
It is NOT possible unless you know the dimensions (either mass or length (or distance to from camera), of one objects and the angle and distance between it and a second object in order to calculate the necessary 3D mapping points to measure other objects in frame.
The reason for this is you are attempting to solve a 3 Dimensional problem based on a 2 Dimensional view.
Now going back to the original posters question, re the lunar lander. If you do not know the gravity constant then the arguement above stands. If you bring in external data such as camera information you can calculate it if the requirements in my addendum are met.
Finally for the original poster. Astronomers measuring an unseen planet orbiting another sun use a method where they measure the 'wobble' of the star, and from that calculate the mass required to induce the amount of wobble in said star. In order to make this measurement the astronomer has to know the distance to the star and the mass of the star reasonably accurately.