Projection-Mapping-Based Object Pointing Using a High-Frame-Rate Camera-Projector System
An informative object pointing method using a spatiotemporal-modulated pattern projection is proposed to recognize and localize pointed objects by using a distantly located high-frame-rate vision system. We developed a prototype for projection-mapping-based object pointing that consists of an AI-camera-enabled projection (AiCP) system used as a transmitter, for informative projection mapping, and an HFR vision system operated as a receiver. The AiCP system detects multiple objects in real time at 30 fps with a CNN-based object detector, and simultaneously encodes and projects the recognition results of the detector as 480-Hz-modulated light patterns on to the objects to be pointed. The multiple 480-fps cameras can directly recognize and track the objects pointed at by the AiCP system without camera calibration or complex recognition methods by decoding the brightness signals of pixels in the images. To demonstrate the eectiveness of our proposed method, several desktop experiments using miniature objects and scenes were conducted under various conditions.
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
Figure 8
Figure 9
Figure 10
Figure 11
Figure 12
Figure 13
Figure 14
Due to technical limitations, full-text HTML conversion of this manuscript could not be completed. However, the manuscript can be downloaded and accessed as a PDF.
This is a list of supplementary files associated with this preprint. Click to download.
Posted 18 Sep, 2020
On 20 Nov, 2020
Received 19 Nov, 2020
On 03 Nov, 2020
Received 16 Oct, 2020
On 03 Oct, 2020
Invitations sent on 30 Sep, 2020
On 15 Sep, 2020
On 14 Sep, 2020
On 14 Sep, 2020
On 14 Sep, 2020
Projection-Mapping-Based Object Pointing Using a High-Frame-Rate Camera-Projector System
Posted 18 Sep, 2020
On 20 Nov, 2020
Received 19 Nov, 2020
On 03 Nov, 2020
Received 16 Oct, 2020
On 03 Oct, 2020
Invitations sent on 30 Sep, 2020
On 15 Sep, 2020
On 14 Sep, 2020
On 14 Sep, 2020
On 14 Sep, 2020
An informative object pointing method using a spatiotemporal-modulated pattern projection is proposed to recognize and localize pointed objects by using a distantly located high-frame-rate vision system. We developed a prototype for projection-mapping-based object pointing that consists of an AI-camera-enabled projection (AiCP) system used as a transmitter, for informative projection mapping, and an HFR vision system operated as a receiver. The AiCP system detects multiple objects in real time at 30 fps with a CNN-based object detector, and simultaneously encodes and projects the recognition results of the detector as 480-Hz-modulated light patterns on to the objects to be pointed. The multiple 480-fps cameras can directly recognize and track the objects pointed at by the AiCP system without camera calibration or complex recognition methods by decoding the brightness signals of pixels in the images. To demonstrate the eectiveness of our proposed method, several desktop experiments using miniature objects and scenes were conducted under various conditions.
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
Figure 8
Figure 9
Figure 10
Figure 11
Figure 12
Figure 13
Figure 14
Due to technical limitations, full-text HTML conversion of this manuscript could not be completed. However, the manuscript can be downloaded and accessed as a PDF.