matlab - Connected component labeling (CCL) for vehicle headlight detection -


I am developing a project to detect vehicle headlights in the night scene. I am working on a demo at MATLAB. First of all I use Gaussian blur to filter grayscale image, then threshold to get binary image. The next step is to identify components connected to them and to identify the bounding box. I have CCL code using two pass 4-connectivity method without using BCLL and it is quite easy. But I think this method is very resource when implementing on FPGA. I am considering other algorithms like Contour Tracing or Single Pass Can you suggest some other CCL methods suitable for FPGA? Depending on the type of image you are using and depending on the output of thresholding, many thanks

< P> Operation, you can bypass CCL and go directly to contour tracing.

I can say that the real challenge is to correctly image the threshold, so that only the headlight (as close to possible) in the binary resultant can be.

For example, if I take this image:

image here Enter the details

and do this sequentially in the following steps: 1. Conversion to grayscale 2. Intermediate filter 3. Threshold of oatsu 4. Morphological Gradient

... me The following output is found:

Enter image details here

I think a simple hoo circle on that picture will be enough.


Comments

Popular posts from this blog

ios - How do I use CFArrayRef in Swift? -

eclipse plugin - Run java code error: Workspace is closed -

c - Error on building source code in VC 6 -