Describe Points in Camera and World Coordinates
A point marked on the parking-lot floor has a different location in the camera image.
Describe Points in Camera and World Coordinates explained visually
A point marked on the parking-lot floor has a different location in the camera image. English captions are optional and start off. The complete transcript follows the written lesson.
What you will learn
A point marked on the parking-lot floor has a different location in the camera image.
A practical question
Our parking-lot camera sees a car at one location in the frame. A survey map gives its location in meters. Those are two different coordinate systems, and calling both points x and y won't make them interchangeable.
The useful idea
Name each frame before measuring anything: world coordinates on the lot, camera coordinates relative to the lens, and image coordinates in pixels. A point can move through those descriptions using measured transformations and a camera projection. The order matters. A pixel alone cannot usually tell us the point's distance from the camera.
Work through the example
Mark a painted corner on the ground and write down where it belongs on the site map and in the image. Then check another corner. If the projected point lands in the wrong bay, inspect the frame definitions and units before adjusting a model to compensate.
Limits and next step
Keep the map, image, and conventions together. Next, we'll examine the camera's internal measurements that turn directions into pixel positions.
Worked example
Write the lot's painted corner in world meters and in image pixels with the frame named beside each.
Quick check: test your understanding
Try it yourself: In the situation described above, explain the method and one reason its output might be misleading.
Check the reasoning
A world point and a camera point live in different coordinate frames; a transform relates them. A pixel coordinate is not a distance in meters without geometry.
What is the main idea behind describe points in camera and world coordinates?
A world point and a camera point live in different coordinate frames; a transform relates them.
How is it used in the example?
Write the lot's painted corner in world meters and in image pixels with the frame named beside each.
What limitation should be checked?
A pixel coordinate is not a distance in meters without geometry.
Further reading
Programming Computer Vision with Python — Jan Erik Solem.
Original ToolsInVision teaching examples. Editorial approach · Suggest a correction
Accessible referenceLesson 17 transcript155 words · Video transcript · Matches the published narration and captionsOpen transcript
Our parking-lot camera sees a car at one location in the frame. A survey map gives its location in meters. Those are two different coordinate systems, and calling both points x and y won't make them interchangeable.
Name each frame before measuring anything: world coordinates on the lot, camera coordinates relative to the lens, and image coordinates in pixels. A point can move through those descriptions using measured transformations and a camera projection. The order matters. A pixel alone cannot usually tell us the point's distance from the camera.
Mark a painted corner on the ground and write down where it belongs on the site map and in the image. Then check another corner. If the projected point lands in the wrong bay, inspect the frame definitions and units before adjusting a model to compensate.
Keep the map, image, and conventions together. Next, we'll examine the camera's internal measurements that turn directions into pixel positions.