Skip to main content
added 292 characters in body
Source Link
user315366
user315366

Motion blur in a single image is caused by the finite speed that the camera shutter can open and close when taking pictures. Any object that moves more than 1 pixel distance during this picture-taking process appears blurred.

If this speed is known, the length of the blur (in pixel units) is proportional to how fast the object is moving, and can be tested from experiments using that particular camera.

Motion from a set of images can be also estimated another way, for instance as follows. If a human is around, mark out the object in both photos. Assuming the camera is still, just check one particular identifiable part of the same object in both photos, and note down its difference in pixel coordinates. The object has moved this pixel distance in the time it takes between frames (eg. 1/30 second or 1/60 second).

If we can now estimate how far the object is from the camera, for instance this particular car is probably 2 meters away judging by its appearance and size, we can get an estimate for the 3D distance it has moved between the two frames.

Hope it helps!

PS. In UFO blurred photo cases, sometimes this kind of reasoning is also done, but in this case it is much harder to estimate how far away the unidentified object is, just by the fact that it is "unidentified" so we don't have any experience to draw from, in contrast to the car scenario.

Motion blur in a single image is caused by the finite speed that the camera shutter can open and close when taking pictures. Any object that moves more than 1 pixel distance during this picture-taking process appears blurred.

If this speed is known, the length of the blur (in pixel units) is proportional to how fast the object is moving, and can be tested from experiments using that particular camera.

Motion from a set of images can be also estimated another way, for instance as follows. If a human is around, mark out the object in both photos. Assuming the camera is still, just check one particular identifiable part of the same object in both photos, and note down its difference in pixel coordinates. The object has moved this pixel distance in the time it takes between frames (eg. 1/30 second or 1/60 second).

If we can now estimate how far the object is from the camera, for instance this particular car is probably 2 meters away judging by its appearance and size, we can get an estimate for the 3D distance it has moved between the two frames.

Hope it helps!

Motion blur in a single image is caused by the finite speed that the camera shutter can open and close when taking pictures. Any object that moves more than 1 pixel distance during this picture-taking process appears blurred.

If this speed is known, the length of the blur (in pixel units) is proportional to how fast the object is moving, and can be tested from experiments using that particular camera.

Motion from a set of images can be also estimated another way, for instance as follows. If a human is around, mark out the object in both photos. Assuming the camera is still, just check one particular identifiable part of the same object in both photos, and note down its difference in pixel coordinates. The object has moved this pixel distance in the time it takes between frames (eg. 1/30 second or 1/60 second).

If we can now estimate how far the object is from the camera, for instance this particular car is probably 2 meters away judging by its appearance and size, we can get an estimate for the 3D distance it has moved between the two frames.

Hope it helps!

PS. In UFO blurred photo cases, sometimes this kind of reasoning is also done, but in this case it is much harder to estimate how far away the unidentified object is, just by the fact that it is "unidentified" so we don't have any experience to draw from, in contrast to the car scenario.

Source Link
user315366
user315366

Motion blur in a single image is caused by the finite speed that the camera shutter can open and close when taking pictures. Any object that moves more than 1 pixel distance during this picture-taking process appears blurred.

If this speed is known, the length of the blur (in pixel units) is proportional to how fast the object is moving, and can be tested from experiments using that particular camera.

Motion from a set of images can be also estimated another way, for instance as follows. If a human is around, mark out the object in both photos. Assuming the camera is still, just check one particular identifiable part of the same object in both photos, and note down its difference in pixel coordinates. The object has moved this pixel distance in the time it takes between frames (eg. 1/30 second or 1/60 second).

If we can now estimate how far the object is from the camera, for instance this particular car is probably 2 meters away judging by its appearance and size, we can get an estimate for the 3D distance it has moved between the two frames.

Hope it helps!