1. Multimodal Templates for Real-Time Detection of Texture-less Objects in Heavily Cluttered Scenes
Stefan Hinterstoisser, Stefan Holzer, Cedric Cagniart, Slobodan Ilic,Kurt Konolige, Nassir Navab, Vincent Lepetit
Department of Computer Science, CAMP, Technische Universit¨at M¨unchen (TUM), Germany WillowGarage, Menlo Park, CA, USA
IEEE International Conference on Computer Vision (ICCV) 2011

2. Outline Goal & Challenges Related Work Modality Extraction
Image Cue
Depth Cue
Similarity Measure
Efficient Computation
Experiments

3. Goal

4. Challenges
Objects under different poses over heavily cluttered background
Online learning
Real-time object learning and detection

5. Related Work
Solving the problem of multi-view 3D object detection has two main categories:
Learning Based Methods
Template Matching
Learning Based Methods:
Require a large amount of training data
Require long offline training phase
Expensive learning for new object

6. Related Work Template Matching: Others:
Better adapted to low textured objects than feature point approaches
Easily update template for new object
Direct matching is inappropriate for real-time.
Others:
Matching in Range Data :
Construct full 3D CAD model of the object

7. Outline Goal & Challenges Related Work Modality Extraction
Image Cue
Depth Cue
Similarity Measure
Efficient Computation
Experiments

8. Modality Extraction-Image Cue
Image gradients are proved to be discriminant and robust to illumination change and noise.
Normalized gradients and not their magnitudes makes the measure robust to contrast changes.
We compute the normalized gradients on each color channel for input RGB color image.
Input image , gradient map at location x:

9. Modality Extraction-Image Cue
Keep only the gradients whose norms are larger than a threshold.
Assign to the gradient whose quantized orientation occurs most in a 3 × 3 neighborhood.
The similarity measurement function fg:
Og(r): the normalized gradient map of the reference image at location r
Ig(t): the normalized gradient map of the input image at location t

10. Modality Extraction-Image Cue
Quantizing the gradient orientations
Gradient image computed on gray image
Gradient image computed with our approach
Input color image

11. Modality Extraction-Depth Cue
We use a standard camera and a aligned depth sensor to obtain depth map.
Use quantized surface normal computed on a dense depth field for our template representation.
Consider the first order Taylor expansion of the depth function D(x):
Within a patch defined around x, each pixel offset dx yields an equation.

12. Modality Extraction-Depth Cue
Estimate an optimal gradient in least-square.
Depth gradient corresponds to a tangent plane going through three points X, X1 and X2:
𝒗 𝒙 : vector along the line of sight that goes through pixel x (obtain from parameters of depth sensor)

13. Modality Extraction-Depth Cue
The normal to the surface can be estimated as the normalized cross-product of X1 − X and X2 − X.
Within a patch defined around x, this would not be robust around occluding contours.
Inspired by bilateral filtering, we ignore the pixels whose depth difference with the central pixel (X) is above a threshold.
D(x)
Tangent
plane 𝑿 𝑿𝟏 𝑿𝟐 X
Depth sensor
Normal of X +Z

14. Modality Extraction-Depth Cue
Quantize the normal directions into n0 bins.
Assign to each location the quantized value that occurs most often in a 5 × 5 neighborhood.
The similarity measurement function fD:
OD(r): the normalized surface normal of the reference image at location r
ID(t): the normalized surface normal of the input image at location t

15. Modality Extraction-Depth Cue
Quantizing the surface normals
Input
image
The corresponding
depth image
Surface normals computed with our approach.
Details are clearly visible and depth discontinuities are well handled.

16. Outline Goal & Challenges Related Work Modality Extraction
Image Cue
Depth Cue
Similarity Measure
Efficient Computation
Experiments

17. Similarity Measure We define a template as T = ({Om} m∈M, P ).
P: a list of pairs (r,m) made of the locations r of a discriminant feature in modality m.
Each template is created by extracting for each “m” a set of its most discriminant features (P).
P:(ri, gradients)
P:(rk, surface normals)
r : record the
feature location
with respect to
object center (C). C

18. Similarity Measure The object measurement energy function :
T: ({Om} m∈M, P )
c: the detected location (could be object center)
R(c+r): [ c+r- 𝑁 2 , c+r+ 𝑁 2 ]×[ c+r- 𝑁 2 , c+r+ 𝑁 2 ] , N ∈ const. (neighborhood of size N centered on (c+r) in Im)
fm (Om (r), Im (t)) : computes the similarity score for modality m

19. Efficient Computation
We first quantize the input data for each modality into a small number of n0.
Use a lookup table тi,m for energy response:
i: the index of the quantized value of modality m. (also use i to represent the corresponding value)
Lm: list of values of a special modality m appearing in a local neighborhood of a value i from input I. Lm
Lm’ C C’

20. Efficient Computation
“Spread” [11] the data around neighborhood to obtain a robust representation Jm instead of Lm.
For each quantized value of one modality m with index i we can now compute the response at each location c:
тi,m : the precomputed lookup table, Jm as the index
[11] S. Hinterstoisser, C. Cagniart, S. Ilic, P. Sturm, P. Fua, N. Navab, and V. Lepetit. Gradient
response maps for realtime detection of texture-less objects. under revision PAMI.

21. Efficient Computation
Finally, the similarity measure can be:
Since the maps Si,m are shared between the templates, matching several templates against the input image can be done very fast once they are computed.

22. Experiments LINE-MOD: our approach (intensity & depth)
LINE-2D: introduced in [11] (use only intensity)
LINE-3D: use only the depth map
Hardware:
Performed on one processor of a standard notebook with an Intel Centrino Processor Core2Duo with 2.4 GHz and 3 GB of RAM.
Test data:
Six object sequences made of 2000 real images each.
Each sequence presents illumination and large viewpoint changes over heavy cluttered background.

23. Experiments Robustness:
A threshold (about 80) separates almost all true positives for LINE-MOD.

24. Experiments
Speed:
Learning new templates only requires extracting and storing features, which is almost instantaneous.
Templates include: 360 degree tilt rotation, 90 degree inclination rotation and in-plane rotations of ± 80 degrees, scale changes from 1.0 to 2.0.
Parse a 640×480 image with over 3000 templates with 126 features at about 10 fps(real-time).
The runtime of LINE-MOD is only dependent on the number of features and independent of the object/template size.

25. Experiments
Speed:

26. Experiments Occlusion:
Right: Average recognition score for the six objects with respect to occlusion.
With over 30% occlusion our method is still able to recognize objects.

27. Experiments
True positive rates
Cup
Toy-Car
Hole punch

28. Experiments
Toy-Monkey
Toy-Duck
Camera

29. Experiments
True positive rates = 𝑇𝑟𝑢𝑒 𝑝𝑜𝑠𝑖𝑡𝑖𝑣𝑒𝑠 𝑇𝑟𝑢𝑒 𝑝𝑜𝑠𝑖𝑡𝑖𝑣𝑒𝑠+𝐹𝑎𝑙𝑠𝑒 𝑛𝑒𝑔𝑎𝑡𝑖𝑣𝑒 ×100%
False positive rates = 𝐹𝑎𝑙𝑠𝑒 𝑝𝑜𝑠𝑖𝑡𝑖𝑣𝑒𝑠 𝑇𝑟𝑢𝑒 𝑛𝑒𝑔𝑎𝑡𝑖𝑣𝑒+𝐹𝑎𝑙𝑠𝑒 𝑝𝑜𝑠𝑖𝑡𝑖𝑣𝑒𝑠 ×100%

30. Experiments