Minimal Geometric Representation and Strawberry Stem Detection

This paper takes a crucial step toward a visual system for an automated strawberry harvester. We present an algorithm based on the Blum medial axis that outputs for a given berry image a bounding box containing the berry's stem, and determines minimal geometric information to do so. The algorithm first generates three potential boxes, then automatically selects which of the three contains the stem. We compare the performance of our geometric-based stem detection with two other methods. The first, implemented already for a berry harvesting robot, relies on the principal axes of the berry shape to define the bounding box. The second takes as input the three potential boxes generated using the medial axis, then selects the one containing the stem by computing geometric and appearance features within each box for use in an ensemble classifier of 250 trees boosted by RUSboost with five-leaf minimum and a learning rate of 0.1. Note that because our data is imbalanced we used class-proportional sampling. Our geometric approach outperforms the other two methods on a database of 286 strawberry images.

[1]  Paul A. Yushkevich,et al.  Continuous medial representations for geometric object modeling in 2D and 3D , 2003, Image Vis. Comput..

[2]  N. Otsu A threshold selection method from gray level histograms , 1979 .

[3]  H. Blum Biological shape and visual science (part I) , 1973 .

[4]  Kathryn Leonard,et al.  An Efficiency Criterion for 2D Shape Model Selection. , 2006, 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'06).

[5]  Kathryn Leonard,et al.  Efficient Shape Modeling: ⋮-Entropy, Adaptive Coding, and Boundary Curves -vs- Blum’s Medial Axis , 2007, International Journal of Computer Vision.

[6]  Keigo Watanabe,et al.  Behavior Selection Based Navigation and Obstacle Avoidance Approach Using Visual and Ultrasonic Sensory Information for Quadruped Robots , 2008 .

[7]  Tamal K. Dey,et al.  Approximating the Medial Axis from the Voronoi Diagram with a Convergence Guarantee , 2003, Algorithmica.

[8]  H. Blum Biological shape and visual science. I. , 1973, Journal of theoretical biology.

[9]  Benjamin B. Kimia,et al.  On the intrinsic reconstruction of shape from its symmetries , 1999, Proceedings. 1999 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (Cat. No PR00149).

[10]  Kokichi Sugihara,et al.  A method for stable construction of medial axes in figures , 2002 .

[11]  J. MacQueen Some methods for classification and analysis of multivariate observations , 1967 .

[12]  Jorma Rissanen,et al.  Stochastic Complexity in Statistical Inquiry , 1989, World Scientific Series in Computer Science.

[13]  James N. Damon,et al.  Determining the Geometry of Boundaries of Objects from Medial Data , 2005, International Journal of Computer Vision.

[14]  Kaleem Siddiqi,et al.  A shock grammar for recognition , 1996, Proceedings CVPR IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[15]  Cao Qixin,et al.  Fruit Detachment and Classification Method for Strawberry Harvesting Robot , 2008 .

[16]  Yael Edan,et al.  Computer vision for fruit harvesting robots - state of the art and challenges ahead , 2012, Int. J. Comput. Vis. Robotics.

[17]  Serge Beucher,et al.  Use of watersheds in contour detection , 1979 .