Computer Vision interview Question Part-2
1 – What programming languages does computer vision support?
Computer vision is a field of computer science and engineering that deals with how computers can be made to gain high-level understanding from digital images or videos.
The following are the programming languages that support computer vision:
– C++
– Java
– Python
– Prolog
2 – How do you project a 3D point to an image?
This is a difficult task for computer vision. To project a 3D point to an image, you need to know the camera’s position and orientation in 3D space. Then, you need to know the camera’s focal length, the distances from the object to both of the cameras, and finally, the coordinates of all points on the object in 3D space that are visible from both cameras.
3 – Explain the “mach band effect.”
The Mach band effect is a visual artifact that occurs when two adjacent surfaces reflect light at different intensities. The effect is typically seen on a straight, parallel edge of an object.
The Mach band effect is caused by the contrast of brightness between the two surfaces. As light hits one surface, it scatters in many directions. Some of this scattered light will hit the second surface and be reflected back to your eye. The intensity at which these reflections hit your eye depends on how far away the second surface is from you, and also on how reflective it is compared to the first surface. If you are close to one side of an object and far from the other side, then you will see a brighter reflection from that closer side and a darker reflection from that farther side.
4 – Define Sampling And Quantization?
Sampling is the process of taking a subset of data and making inferences about the whole dataset. In computer vision, sampling is used to reduce the complexity of a problem by selecting only a few features.
Quantization is the process of reducing numbers to a finite number of discrete values. This process can be done in different ways:
– rounding: rounding off numbers to an integer value
– truncating: chopping off all digits after some point
– discretizing: dividing numbers into groups
5 – Do You Know What Rectification Is In Image Processing?
Rectification is the process of aligning a distorted image so that it can be viewed accurately, and this process is done by computer vision. It involves finding correspondences between points in the input image and points in the ground truth, then computing an affine transformation to map one to the other. The rectification process is typically used for images taken with a wide-angle lens or fish-eye lens because they have such an extreme distortion that they are difficult to view without rectifying them. Rectification has also been used on images taken with fisheye lenses mounted on video cameras in order to produce a more traditional panoramic effect.