CIRP Journal of Manufacturing Science and Technology 14 (2016) 2027

Contents lists available at ScienceDirect

CIRP Journal of Manufacturing Science and Technology

journal homepage: www.elsevier.com/locate/cirpj

A method for detection of randomly placed objects for robotic handling

Panagiota Tsarouchi, Stereos-Alexandros Matthaiakis, George Michalos, Sotiris Makris,
George Chryssolouris *
Laboratory for Manufacturing Systems and Automation, Department of Mechanical Engineering and Aeronautics, University of Patras, Patras 26500, Greece

A R T I C L E I N F O A B S T R A C T

Article history: The image based systems still have open issues in order to meet the latest manufacturing requirements
Available online 23 April 2016 for simplicity, low cost as well as the limited maintenance requirements. In this direction, there is a
method proposed for the recognition of 3D randomly placed objects for eventual robotic handling. The
Keywords: method includes a 2D vision system and is combined with data from computer-aided design (CAD) les
Object recognition for the generation of 3D coordinates. It is generic and can be used for the identication of multiple
Image processing algorithm randomly placed objects. The method has been implemented in a software tool using MATLAB1 and has
Robot been applied to a consumer goods case for the recognition of shaver handles.
Sensor 2016 The Authors. This is an open access article under the CC BY-NC-ND license (http://
Measurement
creativecommons.org/licenses/by-nc-nd/4.0/).

Introduction An extended review on the sensor technologies and the research

Important aspects of vision systems, in robotic applications, are
the simplicity of the algorithm, the low cost and the reduced need for
maintenance, while aspects such as the fast and effective identica-
tion still constitutes an unsolved problem. Even though adequately
efcient and accurate algorithms have been developed, the proces-
sing speed still fails to meet the modern manufacturing requirements
[1]. The problem becomes further complicated owing to the objects
properties, such as shape, material, colour, etc. Additionally, the
requirements for simplicity and low cost are directly connected to the
production rate that is expected to be increased with the introduction
of robotic equipment in modern production lines [2].
The need for objects recognition systems is met in multiple
industrial applications, where different objects of variable shapes
and sizes should be handled. An example of such an application is
the consumer goods industry. This sector lacks in exible, low cost
and simple automation solutions that will eventually cope with
high production rates and a variety of products (Fig. 1(a)).
Currently, the high speed feeding and handling of objects imposes
the need for dedicated equipment, namely feeder bowls. Such
solutions are characterized as noisy, expensive and are dedicated
to product specic equipment. Specically, feeder bowls (Fig. 1(b))
cannot perform in more than one product, whilst for new products
need redesigning.
* Corresponding author. Tel.: +30 2610 997262; fax: +30 2610 997744.
E-mail address: xrisol@otenet.gr (G. Chryssolouris).
trends for assembly systems have shown that vision systems are
suitable for objects recognition, inspection and robot handling
applications [38]. Vision-based robot control is investigated in [9
14], while a survey on the visual servoing systems is presented in
[1517]. Research has been done in the design aspects of the
machine vision systems for industrial applications [18], and has led
to improvements in reliability and product quality [1,19,20]. The
vision system classication was investigated in [2127].
The pattern recognition method has also been investigated in
[14,28,29], while 3D vision systems have been presented in [30
38]. Such systems have the advantages of recognizing the objects
characteristics, but they are based on complicated algorithms and
are prompt to failures in industrial environments. Methods for
errors measurement in vision systems have been researched in
[39], while techniques for the classication of objects and point
descriptors are designated in [4044]. Histogram-based image
descriptors have been evaluated for the classication of 3D objects
[45], while a comparison between the methods of local and full
ranking point descriptors was reported in [46,47].
A low cost system is proposed for 3D randomly placed objects in
order to meet the challenges of the semi-structured industrial
environment. This system is based on simple algorithms for the
recognition of 2D objects combined with CAD data for the
computation of 3D coordinates. The use of CAD data, instead of
a 3D vision system, reduces the complexity and the cost. The object
pose recognition is based on observations with the physical object,
while the measurement of 3rd coordinates, along the z axis, was
made with the help of the object design les (CAD). The techniques

http://dx.doi.org/10.1016/j.cirpj.2016.04.005
1755-5817/ 2016 The Authors. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
 P. Tsarouchi et al. / CIRP Journal of Manufacturing Science and Technology 14 (2016) 2027
Fig. 1. (a) Variety of products in consumer goods industry and (b) feeder bowls.
of the 2D vision system comprise basic image processing
algorithms for the identication of the Points of Interest (POIs)
and the object pose identication in the image. The method is
applied to a consumer goods pilot case.
Method description
The proposed method includes (a) an ofine system using the
CAD les of the objects and (b) an online system for the recognition
of object poses and 2D POIs (Fig. 2). The ofine system is used for
the observation of the candidate objects in order for the different
poses and the POIs to be decided upon. Observations with physical
objects allow the identication of the different objects n poses and
the denition of the POIs. The distance from the objects lying
surface and the POI along the z axis is measured in a CAD software
platform for the n different poses. The z axis coordinate of each
different POI is stored and used later on, in the online vision
system.
The image based recognition system runs online in order to
detect the objects poses and compute the 2D POIs coordinates. The
transformation between the image plane and the world reference
frame (WRF) of the estimated 2D POI follows. Since the 2D POI
coordinates have been estimated, the 3D POI coordinates are
generated with the help of the ofine system.
The online system starts with basic image processing techni-
ques and principles, via the combination of algorithms for object
pose detection and POI recognition. The geometrical features, such
Fig. 2. Method overview.
21
as the pose, orientation, coordinates of the POIs, etc. are computed
via the binary image processing and the contrast value measure-
ments in the blue band of the RGB image. In the RGB image, the
value of 255 is considered being 100% visible in terms of blue
colour and 0 of being fully transparent, based on a single 8-bit byte
system. The objects shape is approximated with an ellipse and
some of its geometrical features are computed with the help of the
ellipse, comprising the major, minor axis, edges, middle points,
orientation, etc. Other features that have been taken into account
are the centroid and middle points, which are estimated by diving
the distance of the major axis.
The next step concerns the automatic objects pose detection.
This method is based on the maximum distance between the k
middle points and the centroid point. The poses are decided upon
the distances of these k points (Eq. (3.1)), as well as their contrast
values in the blue band. The POI automatic recognition is
computed on the basis of the highest contrast values, compared
with those of the middle points, depending on the object.
k
DIFF k jCMP C CP j (3.1)
where:
 DIFFk is the absolute difference of the contrast values between
the k middle points and the centroid;
k
 CMP is the contrast value of the kth middle point;
 CCP is the contrast value of the centroid point.
22 P. Tsarouchi et al. / CIRP Journal of Manufacturing Science and Technology 14 (2016) 2027

The generation of the POIs 3D coordinates in the WRF  The proposed method for the detection of the object pose is
comprises the nal target of the online system. The reference based on the measurement of the measurement of the points
systems transformations into the WRF follow. Two transforma- contrast values, as well as on the computation of the objects
tions are used among the three different reference frames, namely intermediate points;
the camera, the pixel and the world reference frame.  The method for the 2D POI recognition is based on the absolute
The rst transformation is estimated with the following differences of the contract values between the different points on
equation: the object.
2 3 2 32 3
xim 1=sx 0 ccx x
4 yim 5 4 0 1=sy ccy 54 y 5 (3.2)
1 0 0 1 1 System implementation

where: The proposed system has been implemented as a software tool

in MATLAB1 (Fig. 3). Algorithms for the detection of the image
 xim, yim are the coordinates on the image plane; processing, objects poses, calculation and transformation of the 2D
 (ccx, ccy) are the coordinates of the principal point (cc) on the POIs coordinates have been developed under this tool. The results
image plane reference frame; of the ofine system have been stored into a text le and are used
 sx and sy are the effective size of the pixels in the horizontal and as an input in the proposed online system.
vertical directions, estimated in millimetres. A 2D vision sensor has been used for the acquisition of images.
This 2D vision sensor is a high resolution 2MP Basler A641FC camera,
The point in the pixels frame is the vector [x; y; 1] which denotes the which is equipped with 1/1.8 CCD colour image sensor, at a
transformed coordinates of the POI, with respect to the image plane maximum resolution of 1624  1234 pixels. This camera is mounted
reference system. The POI coordinates are estimated by Eq. (3.3). on a xed distance from the conveyor, namely Zc = 415 mm. The
2 3 2 31 2 3 cameras focal length is 7.1 mm and is equipped with a Computar
x 1=sx 0 ccx xim M0814-MP2 lens, which supports the manual iris and focus control.
4y5 4 0 1=sy ccy 5 4 yim 5 (3.3)
1 0 0 1 1
Pilot case study
The principal points projection on the surface that the object is
lying on, considering the cameras pinhole model, is the WRF. Any Description and layout setup
dened user frame can be used as a WRF, with the application of
the appropriate transformation. The transformation that provides The proposed method has been applied to a consumer goods
the WRF coordinates, with respect to the image plane reference industry case study in order to replace the manual or semi-
system, is presented in the following equation: automated ways for the feeding and packaging of handles
2 3 2 32 3 (Fig. 4(a)). The setup includes two industrial robots placed
xWRF Zc=f 0 0 x
4 yWRF 5 4 0 symmetrically in front of a moving conveyor that is fed with
Zc=f 0 54 y 5 (3.4)
zWRF 0 0 Zc 1 shaver handles. An industrial camera is placed at the forepart of the
conveyor belt, for the recognition of the randomly placed shaver
where: handles (Fig. 4(b)).
The vision system may provide a reliable solution for cost
 Zc is the distance from the horizontal measurement surface to effective recognition of multiple objects and thus may enable high
the cameras sensor surface; speed handling by articulated. Cost efciency will be achieved by:
 f is the cameras focal length.
 The use of 2D combined with CAD data les, instead of 3D vision
The 3D POI coordinates are computed as the result of the image system;
based recognition and the measurements on the CAD les, given  The replacement of the heavy feeder bowls with the proposed
the different objects poses. The unique features of this research are system;
summarized as follows:  The re-conguration of the cell for similar products allowed by
the robotized solution.
 The combination of an ofine system that is based on the poses
identication and the 3rd coordinate estimation, through CAD The high production rate of 100 objects/min imposes the need
les measurements and an online 2D vision system for 3D for simple and fast object detection and pose estimation and
coordinates computation; therefore, the use of a slow 3D scanning system is not viable.

Fig. 3. System implementation overview.

 P. Tsarouchi et al. / CIRP Journal of Manufacturing Science and Technology 14 (2016) 2027
Fig. 4. (a) Consumer goods industry scenario and (b) system setup.
Fig. 5. (a) Shaver handles geometry and (b) different poses of the object.
Instead, the use of the objects CAD les enables the supplementa-
tion of the measurement by the 2D system with information,
regarding the 3rd coordinate of the identied POIs thus, minimiz-
ing the computational time.
Application of the method in the shaver handles recognition
In the case of consumer goods, the object is the shavers handle.
The POI as well as the centroid and middle points are illustrated in
Fig. 5(a). Based on the geometry of the shavers handle, which is
randomly dropped on the at belt, as shown in Fig. 5(b), three
different poses are observed. The shavers handle is found in three
poses, from left to right: in the back up (pose 1), lying on its side
(pose 2) and the back down pose (pose 3). Depending on these
poses, three different coordinates of the z axis are considered from
the CAD les.
The proposed online system workow, based on the informa-
tion identied by the three different poses, is applied to the
consumer goods industry, as illustrated in Fig. 6.
The transformation between the camera reference frame and
the WRF in the case of the shaver handle is illustrated in Fig. 7.
Results of 2D vision system recognition
The proposed method has been successfully tested on numer-
ous images with different object orientation and poses. The
obtained results include the assignment number of the recognized
object, the orientation with respect to the x axis of the image plane,
the pose of the object and the 3D coordinates of the POI with the
23
help of a vision system and the CAD les. Some examples of
recognition results are illustrated in Fig. 8. Four different images
and their recognition results, along with the RGB images in the
corresponding binary ones, are presented in this gure.
The results of the four images with IDs 1, 2, 3, 6 as well as three
additional ones are reported in Table 1. For each image, there is a
presentation of the number of objects, the pose of shaver handles,
the 3D POI coordinates, as well as their orientation.
The results in an RGB image, with more than three shaver
handles, are visualized in Fig. 9. Nine shaver handles were found in
this image and their features were successfully recognized.
An example of the proposed online system of shaver handles, is
illustrated, in more detail, in Fig. 10. The ellipse shape was adjusted
to every handle that was found in the image. The major axis of each
ellipse has shown each objects orientation in respect to the x axis
of the image plane, where the centroid and middle points were
estimated. Additionally, the POI is automatically computed and
illustrated in yellow colour. The parallel lines to the minor axis of
ellipse were used for the estimation of the sum of all the pixels
contrast values, in order for the objects pose to be decided upon
according to the colour of each one of the poses.
Comparison with other methods
Different methods were also evaluated for the image based
recognition system. The point descriptors, is one of these methods
for the evaluation of the Speeded-Up Robust Features (SURF)
features extraction. An example of the SURF features, recognized in
the shaver handles images, is visualized in Fig. 11, where the
24 P. Tsarouchi et al. / CIRP Journal of Manufacturing Science and Technology 14 (2016) 2027

Fig. 6. Flowchart of online system for shaver handles recognition.

Fig. 7. Transformation from camera to WRF for a shaver handle.

function detectSURFFeatures is used in the MATLAB1 toolbox.
These images are the same as those presented in Fig. 8. This
function implements the Speeded-Up Robust Features (SURF)
algorithm, described in [47] and nds blob features. The detection
of features is useful when they represent specic data structures in
the image, such as specic points, edges, corners, etc. The SURF
detected features were not enough for the recognition of all the
different poses of the shaver handles; as regards the same objects
and the same pose, there were differences in the number of
detected features.
The fact that the object was symmetric and that some pairs of
the point descriptors confused the algorithm for the object
 P. Tsarouchi et al. / CIRP Journal of Manufacturing Science and Technology 14 (2016) 2027 25

Table 1
Method results for the shaver handles.

Image Object Pose 3D POI coordinates Orientation (8)

ID ID [x, y, z] (mm)

1 1 Lying on [154.919, 45.048, 11.040] 64.939

its side
1 2 Back down [80.057, 2.386, 22.450] 85.993
1 3 Back up [11.925, 99.667, 1.240] 58.514
2 1 Back up [93.292, 96.309, 1.240] 101.725
2 2 Back up [61.417, 27.749, 1.240] 98.733
2 3 Back up [26.865, 124.338, 1.240] 69.748
3 1 Back up [76.984, 1.854, 1.240] 38.438
3 2 Lying on [66.157, 88.544, 11.040] 77.0185
its side
3 3 Back up [39.118, 101.755, 1.240] 69.556
4 1 Back down [108.925, 18.749, 22.450] 80.666
4 2 Back down [38.017, 0.972, 22.450] 60.1968
4 3 Back down [6.336, 126.759, 22.450] 87.8413
5 1 Lying on [159.301, 111.633, 11.040] 70.867
its side Fig. 10. Image based object recognition an example.
5 2 Lying on [86.809, 0.785, 11.040] 113.306
its side
5 3 Lying on [3.120, 123.931, 11.040] 88.253 recognition, did not help the evaluation of this method. As a result,
its side
this method could not be taken into consideration for the specic
6 1 Back down [122.659, 114.977, 22.450] 78.387
6 2 Lying on [43.203, 13.945, 11.040] 74.715
objects, on its own, since the features that were recognized did not
its side help in the direction of the POIs computation. The direction of
6 3 Back up [43.298, 119.217, 1.240] 42.333 combining multiple algorithms for features extraction, might have
7 1 Lying on [24.633, 95.057, 11.040] 115.024 had better results, as the multi-level feature processing could
its side
possibly provide a solution to the POIs computation. This
7 2 Back down [98.210, 5.347, 22.450] 100.577
7 3 Back down [15.460, 116.101, 22.450] 128.185 consideration is oriented towards the specic methodologys

Fig. 8. RGB images and recognition results in binary images.

Fig. 9. (a) RGB image with multiple objects and (b) recognition results.
26 P. Tsarouchi et al. / CIRP Journal of Manufacturing Science and Technology 14 (2016) 2027
Fig. 11. SURF features extraction.
application and would not help if it were applied to similar objects
without any change on the feature processing strategy.
Accuracy and errors
The POIs x and y axis coordinates were compared with the real
point coordinates, since they were measured with the help of a
millimetre paper, placed on the working area. Forty-eight different
x and y axis coordinated images were compared to their real
position according to this WRF. The millimetre paper was tted
into the working area and the real positions in the eld of view,
with a pencil and ruler, were measured in order for the POIs of each
shaver handle to be measured.
The average relative errors in the two axes are 2.00355% in x and
3.49674% in the y axis respectively. It was observed that when the
POI was found near the cc projection area, the relative error was
minor, in contrast to that when the POIs were found far away from
this point. From an applications point of view, these errors in a
future study will not have a signicant effect on the gripping
devices ability to accurately approach, pick up and nally handle
the shavers.
Discussion
The proposed study combines the advantages of an image based
system and the use of data from CAD for the recognition of 3D
objects and the estimation of coordinates. Currently, a minor issue
on the proposed system is the use of contrast values in order for the
measurements to be performed, thus having a signicant increase
in the computation time. Despite that fact, this method helps with
the designing of similar vision systems in different objects, by
adjusting the computational logic of contrast values. In this way,
the proposed method can be generalized in a variety of small 3D
objects, indifferent shapes, for similar applications. Additionally, it
is quite exible and allows the recognition of randomly and not
well-oriented objects in the working area, similar to other 3D
vision system applications, such as bin picking.
Some examples of commercial tools for such applications are
among others the ISRA vision system [48], the Visio Nerf [49], the
FANUC iR PickTool [50] and the 3D Area Sensor [51], as well as the
Pick-itTM Robotic Picking [52]. Despite these tools being commer-
cial and performing well in such applications, they entail
challenges, namely the high cost, the vision sensor dimensions
and weight, as well as the use of laser scanners and lighting
mechanisms, all of which make their application unaffordable.
Additionally, further challenges are directly related to the integra-
tion aspects with the robot control systems. Given these challenges,
this study has come up with a solution that is based on an image
based method for 3D objects recognition using CAD les data.
Conclusions and future work
The development of functions for the POI recognition, in various
lighting conditions, was based on the combination of the objects
geometrical features and the contrast values. The recognition
result was completed with the use of the objects CAD data. The
benets of the proposed system among others include:
 The estimation of the relative error being 2.00355% in x and
3.49674% in the y axis respectively; comparatively it is quite
good for similar applications, from an accuracy point of view.
 The fact that an online 2D vision system was combined with
ofine data from CAD les for the computation of 3D POI
coordinates allows for a signicant decrease in the cost of the
nal system, as well as a decrease in the setup time.
Future study will be focusing on this systems implementation and
improvement in the cell setup, along with the robots and the
mechatronics systems involved. The online system will be appropri-
ately modied in order for the POIs to be recognized and thus enable
the completion of the handling process by the gripping devices, on a
moving conveyor. The challenges met in this kind of vision systems
not only do they concern the recognition and computation of the 3D
coordinates, but also the consideration of real life problems. Such
problems include the existence of dust, the need for extra lighting,
distortions, etc. which are common in industrial facilities as well as the
effect of shiny surfaces that cause light reections and even
occlusions, such as those on the moving conveyor.
Experiments will be conducted to test the accuracy of the 3D
coordinates calculation and the way that the data of the robots
TCP (Tool Center Point) will be managed for its synchronization
with the moving conveyor. Last but not least, the design and
development of a gripper for the handling of such objects has been
considered for future work. The experiments performed, under the
AUTORECON FP7 EU project, aim at providing a robust and efcient
vision system, which could guide both robots to handle the object
at a rate of 100 objects/min.
Acknowledgements
This paper is an extension of the conference paper presented at
the 2nd CIRP Global Web Conference, CIRPe2013, held on 1112
June 2013 [53]. The research has been partially supported by the
research project AUTORECON Autonomous co-operative
machines for highly recongurable assembly operations of the
future funded by the European Commission.
References
[1] Kodagali, J., Balaji, S., 2012, Computer Vision and Image Analysis Based
Techniques for Automatic Characterization of Fruits A Review, International
Journal of Computer Applications, 50/6: 612.
[2] Chryssolouris, G., 2006, Manufacturing Systems: Theory and Practice, Springer.
[3] Santochi, M., Dini, G., 1998, Sensor Technology in Assembly Systems, CIRP
Annals Manufacturing Technology, 503524. http://dx.doi.org/10.1016/
S0007-8506(07)63239-9.
[4] Tonshoff, H.K., Janocha, H., Seidel, M., 1988, Image Processing in a Production
Environment, CIRP Annals Manufacturing Technology, 579590. http://
dx.doi.org/10.1016/S0007-8506(07)60755-0.
 P. Tsarouchi et al. / CIRP Journal of Manufacturing Science and Technology 14 (2016) 2027
[5] Michalos, G., Makris, S., Eytan, A., Matthaiakis, S., Chryssolouris, G., 2012, Robot
Path Correction Using Stereo Vision System, Procedia CIRP, 352357. http://
dx.doi.org/10.1016/j.procir.2012.07.061.
[6] Vahrenkamp, N., Boge, C., Welke, K., Asfour, T., Walter, J., et al, 2009, Visual
Servoing for Dual Arm Motions on a Humanoid Robot, 9th IEEE-RAS Interna-
tional Conference on Humanoid Robots, HUMANOIDS09, pp.208214. http://
dx.doi.org/10.1109/ICHR.2009.5379577.
[7] Han, S.H., See, W.H., Lee, J., Lee, M.H., Hashimoto, H., 2000, Image-Based Visual
Servoing Control of a SCARA Type Dual-Arm Robot, ISIE2000. Proceedings of
the 2000 IEEE International Symposium on Industrial Electronics (Cat. No.
00TH8543), vol. 2, . http://dx.doi.org/10.1109/ISIE.2000.930351.
[8] Han, S.H., Choi, J.W., Son, K., Lee, M.C., Lee, J.M., et al, 2001, A Study on Feature-
Based Visual Servoing Control of Eight Axes-Dual Arm Robot by Utilizing
Redundant Feature, in: Proceedings 2001 ICRA. IEEE International Conference
on Robotics and Automation (Cat. No.01CH37164), vol. 3, . http://dx.doi.org/
10.1109/ROBOT.2001.933018.
[9] Hutchinson, S., Hager, G.D., Corke, P.I., 1996, A Tutorial on Visual Servo Control,
IEEE Transactions on Robotics and Automation, 12/5: 651670. http://
dx.doi.org/10.1109/70.538972.
[10] Chaumette, F., Hutchinson, S., 2006, Visual Servo Control. I. Basic Approaches,
IEEE Robotics and Automation Magazine, 13/4: 8290. http://dx.doi.org/
10.1109/MRA.2006.250573.
[11] Chaumette, F., Hutchinson, S., 2007, Visual Servo Control. II. Advanced
Approaches [Tutorial], IEEE Robotics and Automation Magazine, 14/1: 109
118. http://dx.doi.org/10.1109/MRA.2007.339609.
[12] Makris, S., Tsarouchi, P., Surdilovic, D., Kruger, J., 2014, Intuitive Dual Arm
Robot Programming for Assembly Operations, CIRP Annals Manufacturing
Technology, 63/1: 1316. http://dx.doi.org/10.1016/j.cirp.2014.03.017.
[13] Zhao, Y., Cheah, C.C., 2009, Neural Network Control of Multingered Robot
Hands Using Visual Feedback, IEEE Transactions on Neural Networks, 20/5:
758767. http://dx.doi.org/10.1109/TNN.2008.2012127.
[14] Lippiello, V., Siciliano, B., Villani, L., 2005, An Experimental Setup for Visual
Servoing Applications on an Industrial Robotic Cell, in: Proceedings, 2005 IEEE/
ASME International Conference on Advanced Intelligent Mechatronics, . http://
dx.doi.org/10.1109/AIM.2005.1511212.
[15] Kragic, D., Christensen, H.I., 2002, Survey on Visual Servoing for Manipulation,
Report, ISRN KTH/NA/P-02/01-SE, .
[16] Malis, E., 2002, Survey of Vision-Based Robot Control, European Naval Ship
Design, Captain Computer IV Forum, ENSIETA, Brest, France.
[17] Kelly, R., 1996, Robust Asymptotically Stable Visual Servoing of Planar Robots,
IEEE Transactions on Robotics and Automation, 12/5: 759766. http://
dx.doi.org/10.1109/70.538980.
[18] Golnabi, H., Asadpour, A., 2007, Design and Application of Industrial Machine
Vision Systems, Robotics and Computer-Integrated Manufacturing, 23/6: 630
637. http://dx.doi.org/10.1016/j.rcim.2007.02.005.
[19] Beserra Gomes, R., Ferreira da Silva, B.M., Rocha, L.K.D.M., Aroca, R.V., Velho,
L.C.P.R., et al, 2013, Efcient 3D Object Recognition Using Foveated Point
Clouds, Computers & Graphics, 37/5: 496508. http://dx.doi.org/10.1016/
j.cag.2013.03.005.
[20] Xie, S.Q., Haemmerle, E., Cheng, Y., Gamage, P., 2008, Vision-Guided Robot
Control for 3D Object Recognition and Manipulation, in Ceccarelli M, (Ed.)
Robot Manipulators. InTech.
[21] Smith, C., Karayiannidis, Y., Nalpantidis, L., Gratal, X., Qi, P., et al, 2012, Dual
Arm Manipulation A Survey, Robotics and Autonomous Systems, 1340
1353. http://dx.doi.org/10.1016/j.robot.2012.07.005.
[22] Han, S.H., See, W.H., Lee, J., Lee, M.H., Hashimoto, H., 2000, Image-Based Visual
Servoing Control of a SCARA Type Dual-Arm Robot, ISIE2000. Proceedings of
the 2000 IEEE International Symposium on Industrial Electronics (Cat.
No.00TH8543), vol. 2, . http://dx.doi.org/10.1109/ISIE.2000.930351.
[23] Kelly, R., Carelli, R., Nasisi, O., Kuchen, B., Reyes, F., 2000, Stable Visual Servoing
of Camera-in-Hand Robotic Systems, IEEE/ASME Transactions on Mechatro-
nics, 5/1: 3948. http://dx.doi.org/10.1109/3516.828588.
[24] Zereik, E., Sorbara, A., Casalino, G., Didot, F., 2009, Autonomous Dual-Arm
Mobile Manipulator Crew Assistant for Surface Operations: Force/Vision-
Guided Grasping, RAST 2009 Proceedings of 4th International Conference
on Recent Advances Space Technologies, pp.710715. http://dx.doi.org/
10.1109/RAST.2009.5158284.
[25] Lippiello, V., Villani, L., Siciliano, B., 2007, An Open Architecture for Sensory
Feedback Control of a Dual-Arm Industrial Robotic Cell, Industrial Robot: An
International Journal, 4653. http://dx.doi.org/10.1108/01439910710718441.
[26] Iwata, H., Sugano, S., 2009, Design of Human Symbiotic Robot TWENDY-ONE,
in: Proceedings IEEE International Conference on Robotics and Automation,
pp.580586. http://dx.doi.org/10.1109/ROBOT.2009.5152702.
[27] Zhou, J., Yu, Y., 2011, Coordination Control of Dual-Arm Modular Robot Based
on Position Feedback Using Optotrak3020, Industrial Robot: An International
Journal, 172185. http://dx.doi.org/10.1108/01439911111106381.
27
[28] Wang, F.-Y., Lever, P.J.A., Pu, B., 1993, A Robotic Vision System for Object
Identication and Manipulation Using Synergetic Pattern Recognition, Robot-
ics and Computer-Integrated Manufacturing, 445459. http://dx.doi.org/
10.1016/0736-5845(93)90008-8.
[29] Richards, C.A., Papanikolopoulos, N.P., 1997, Detection and Tracking for Ro-
botic Visual Servoing Systems, Robotics and Computer-Integrated
Manufacturing, 101120. http://dx.doi.org/10.1016/S0736-5845(96)00034-8.
[30] Kruger, J., Nickolay, B., Heyer, P., Seliger, G., 2005, Image Based 3D Surveillance
for exible Man-Robot-Cooperation, CIRP Annals Manufacturing Technolo-
gy, 1922. http://dx.doi.org/10.1016/S0007-8506(07)60040-7.
[31] Reinhart, G., Tekouo, W., 2009, Automatic Programming of Robot-Mounted 3D
Optical Scanning Devices to Easily Measure Parts in High-Variant Assembly,
CIRP Annals Manufacturing Technology, 58/1: 2528. http://dx.doi.org/
10.1016/j.cirp.2009.03.125.
[32] Lanzetta, M., Santochi, M., Tantussi, G., 1999, Computer-Aided Visual Inspec-
tion in Assembly, CIRP Annals Manufacturing Technology, 1316. http://
dx.doi.org/10.1016/S0007-8506(07)63121-7.
[33] Kruger, J., Lien, T.K., Verl, A., 2009, Cooperation of Human and Machines in
Assembly Lines, CIRP Annals Manufacturing Technology, 58/2: 628646.
http://dx.doi.org/10.1016/j.cirp.2009.09.009.
[34] Reinhart, G., Werner, J., 2007, Flexible Automation for the Assembly in Motion,
CIRP Annals Manufacturing Technology, 56/1: 2528. http://dx.doi.org/
10.1016/j.cirp.2007.05.008.
[35] Reinhart, G., Clarke, S., Petzold, B., Schilp, J., Milberg, J., 2004, Telepresence as a
Solution to Manual Micro-Assembly, CIRP Annals Manufacturing Technolo-
gy, 2124. http://dx.doi.org/10.1016/S0007-8506(07)60636-2.
[36] Pham, T.V., Smeulders, A.W.M., 2005, Object Recognition with Uncertain
Geometry and Uncertain Part Detection, Computer Vision and Image Under-
standing, 99/2: 241258. http://dx.doi.org/10.1016/j.cviu.2005.01.006.
[37] Bai, X., Yang, X., Latecki, L.J., 2008, Detection and Recognition of Contour Parts
Based on Shape Similarity, Pattern Recognition, 41/7: 21892199. http://
dx.doi.org/10.1016/j.patcog.2007.12.016.
[38] Alaoui Mhamdi, M.A., Ziou, D., 2014, A Local Approach for 3D Object Recogni-
tion Through a Set of Size Functions, Image and Vision Computing, 32/12:
10301044. http://dx.doi.org/10.1016/j.imavis.2014.08.015.
[39] Zhu, W., Mei, B., Yan, G., Ke, Y., 2014, Measurement Error Analysis and
Accuracy Enhancement of 2D Vision System for Robotic Drilling, Robotics
and Computer-Integrated Manufacturing, 30/2: 160171. http://dx.doi.org/
10.1016/j.rcim.2013.09.014.
[40] Bicego, M., Castellani, U., Murino, V., 2005, A Hidden Markov Model Approach
for Appearance-Based 3D Object Recognition, Pattern Recognition Letters, 26/
16: 25882599. http://dx.doi.org/10.1016/j.patrec.2005.06.005.
[41] Schindler, K., Suter, D., 2008, Object Detection by Global Contour Shape,
Pattern Recognition, 41/12: 37363748. http://dx.doi.org/10.1016/j.pat-
cog.2008.05.025.
[42] Melkemi, M., Djebali, M., 2001, Weighted A-Shape: A Descriptor of the Shape
of a Point Set, Pattern Recognition, 34:11591170.
[43] Scovanner, P., Ali, S., Shah, M., 2007, A 3-Dimensional SIFT Descriptor and its
Application to Action Recognition, in: Proceedings of the 15th International
Conference on Multimedia MULTIMEDI07, 357. http://dx.doi.org/10.1145/
1291233.1291311.
[44] Frome, A., Huber, D., Kolluri, R., Thomas, B., 2004, Recognizing Objects in Range
Data Using Regional Point Descriptors, European Conference on Computer
Vision, 1:114.
[45] Laptev, I., 2009, Improving Object Detection with Boosted Histograms, Image
and Vision Computing, 27/5: 535544. http://dx.doi.org/10.1016/j.ima-
vis.2008.08.010.
[46] Chan, C.H., Yan, F., Kittler, J., Mikolajczyk, K., 2015, Full Ranking as Local
Descriptor for Visual Recognition: A Comparison of Distance Metrics on Sn,
Pattern Recognition, 48/4: 13281336. http://dx.doi.org/10.1016/j.pat-
cog.2014.10.010.
[47] Bay, H., Ess, A., Tuytelaars, T., Van Gool, L., 2008, Speeded-Up Robust Features
(SURF), Computer Vision and Image Understanding, 110/3: 346359. http://
dx.doi.org/10.1016/j.cviu.2007.09.014.
[48] ISRA Vision, (2015), http://www.isravision.com/.
[49] VISION NERF, (2015), http://www.visionerf.com/fr/.
[50] Vision Functions for Robots, (2015), http://www.fanuc.eu/es/en/robots/
accessories/vision.
[51] Robotic Bin Picking Made Possible with the New FANUC iRVision 3D Area
Sensor, (2015), http://robot.fanucamerica.com/robot-applications/
Bin-Picking-Robot-with-3D-Area-Sensor-Vision-System.aspx.
[52] Pick-itTM Robotic Picking, (2015), http://www.intermodalics.eu/pick-it-3d.
[53] Tsarouchi, P., Michalos, G., Makris, S., Chryssolouris, G., 2013, Vision System for
Robotic Handling of Randomly Placed Objects, Procedia CIRP, 6166. http://
dx.doi.org/10.1016/j.procir.2013.06.169.