In this paper, we propose Button Filtering which is a method of non-linear filtering that creates a set of points from an image. This set of points can be used as coordinates of drones. Button filtering is considered nonlinear filtering because it is not applied as convolution, which is used on typical filtering. Nevertheless, it is easy and fast owing to use of a simple button mask, and it has the advantage of immense extendibility. To date, it is only available for the 8-bit grayscale image, which is generally the edge of the original image or masked image. However, it can be applied to a color image, two-dimensional images in a 3D space, and even three-dimensional objects. In this paper, we present the use of button filtering for any object in any image. To this end, we must perform pre-processing before button filtering. We thus present six methods that can handle almost every image type and identify the specific methods that are effective for certain image types.
[1]
John F. Canny,et al.
A Computational Approach to Edge Detection
,
1986,
IEEE Transactions on Pattern Analysis and Machine Intelligence.
[2]
Javier Alonso-Mora,et al.
Multi-robot navigation in formation via sequential convex programming
,
2015,
2015 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS).
[3]
Hyunchul Shim,et al.
Development of Mini Quadrotor Platform and Experiments on Outdoor Autonomous Swarming Flight
,
2017
.
[4]
Pascal Fua,et al.
SLIC Superpixels Compared to State-of-the-Art Superpixel Methods
,
2012,
IEEE Transactions on Pattern Analysis and Machine Intelligence.
[5]
Andrew Blake,et al.
"GrabCut"
,
2004,
ACM Trans. Graph..
[6]
Nicholas Roy,et al.
Towards A Swarm of Agile Micro Quadrotors
,
2013
.