Description: Target detection and recognition
1 color detection
Collect a lot of image data of the enemy robot, and training, get the color range of the other robot,
And on the threshold of the color detection on the image, to find the suspected enemy robot area quantization
Cheng two values.
2 filter out noise points
The two valued image is processed to remove the noise points in the color detection results.
3 connected region detection
In the image of the suspected region of the connected region detection, calculation of each suspected region of the external contour,
Representation of a rectangular approximation.
4 connected region merging
According to the distance between the connected regions and the color similarity, the distance is close to and the similarity is high.
To merge.
5 shape and size filtering
For a large number of images of the enemy robot training, get the other side of the robot shape information (such as the length and width ratio)
And the size of information (ar
To Search:
File list (Check if you may need any files):
vedio
.....\1.jpg
.....\2.jpg
.....\ACom.cpp
.....\ACom.h
.....\Angle
.....\CameraDS.cpp
.....\CameraDS.h
.....\Coder.txt
.....\Com.cpp
.....\cv - 鍓
Download