SentiSight SDK

Object recognition for robotics and computer vision

SentiSight is intended for developers who want to use computer vision-based object recognition in their applications. Through manual or fully automatic object learning it enables searching for learned objects in images from almost any camera, webcam, still picture or live video in an easy, yet versatile, way.

Available as a software development kit that provides for the development of object recognition systems for Microsoft Windows or Linux platforms.

Reliability and Performance Tests

All tests were performed on an Intel Core i7-2600 processor running at 3.4 GHz.

The SentiSight 3.4 algorithm was tested with a subset of Amsterdam Library of Object Images (ALOI).

  • The subset contained objects 1-100 from ALOI.
  • Images with object viewpoint variations (ALOI-VIEW collection) were used. 36 images per object were used.

The blob- and shape-based algorithms from SentiSight 3.4 were tested separately.

SentiSight 3.4 performance was tested on these image resolutions:

  • 768 x 576 pixels – the original full resolution images from ALOI.
  • 320 x 240 pixels – obtained by resizing the 768 x 576 images before testing.

At 0.1 % False Acceptance Rate (FAR), the recognition rate is from 70 % - 99 %, depending on object structural appearance, transparency, etc. For objects with well-defined internal structure, the recognition rate is 98 % - 99 % at 0.1 % FAR.

SentiSight 3.4 blob-based recognition algorithm - testing
  Without color
usage mode
With color
usage mode
768 x 576 320 x 240 768 x 576 320 x 240
Average learning time
for 1 image
(seconds)
1 thread 0.0638 0.0131 0.0810 0.0159
8 threads 0.0510 0.0111 0.0626 0.0131
Average learning time
for 1 object (36 images)
(seconds)
1 thread 2.2978 0.4710 2.9157 0.5721
8 threads 1.8345 0.3994 2.2523 0.4724
Average recognition speed
(templates per second)
1 thread 30859 163639 15850 92463
8 threads 64775 293318 29492 146647
Average object model size
(kilobytes)
763.42 214.30 814.78 250.85

SentiSight 3.4 shape recognition algorithm - testing
  768 x 576 320 x 240
Average learning time
for 1 image
(seconds)
1 thread 1.0549 0.2149
8 threads 1.0590 0.2148
Average learning time
for 1 object (36 images)
(seconds)
1 thread 37.9769 7.7348
8 threads 38.1227 7.7329
Average recognition speed
(templates per second)
1 thread 185 3650
8 threads 586 8289
Average object model size
(kilobytes)
3313.02 521.64