I'm curious if anybody here has done any work with vanishing point detection on images in Labview. Or i'd like to know of people have recomendations for specific vanishing point detection methods.
I'd like to use this for robot navigation similar to this video: http://www.youtube.com/watch?v=nb0VpSYtJ_Y It looks like the steps for vanishing point detection would be to detect straight edges in the image, superimpose vectors on top the the edges, then find the intersection of those vectors.
Could you give me a little more information about your application and your system as a whole? Are you planning on having a target that you focus towards or a lane that you drive through or a horizon to point at? The Vision Development Module will allow you to do this sort of image acquisition and manipulation including filtering your images and identifying the vanishing point. Obviously, there are numerous ways that you can achieve this and most of them are dependent on the environment that you are operating in. You will also have to create some sort of algorithm to work with your environment including vectors as you mentioned. I hope this helps!
I'm currently attepting to discover the vanishing point on a variable angle camera in a similar situation to the OP.
The samples in the vision library appear to be very basic. Do you know if there are any sample Vi's that you would recommend?
This webpage from Drexel University explains one approach to vanishing point detection that could work with the NI Vision Development Module. The basic approach is to use edge detection to find the lines and then to calculate the intersection of the lines.
I suggest that you start with the Straight Edge Detection example VI that is provided with installation of NI Vision Development Module. In this example, you will see how the software automatically finds the edges of some match sticks. The software could be modified to return the position and angle information about the detected edges. You could use this information to find the projected intersection of the lines in order to estimate the vanishing point.