Implementation of 3D Hand Gesture Recognition System using FPGA

Tsai, Tsung‐Han; Ho, Yuan-Chen; Tsai, Yih-Ru

doi:10.1109/isocc.2018.8649939

Cited by 4 publications

(7 citation statements)

References 3 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A comparison with different studies is listed in Table 1 . Our design of the FPGA is synthesized using the same modules from our previous work [ 10 ]. Compared with [ 10 ], we changed the data flow to reduce the burden of accessing SRAM and optimized the data arrangement.…”

Section: Resultsmentioning

confidence: 99%

“…Our design of the FPGA is synthesized using the same modules from our previous work [ 10 ]. Compared with [ 10 ], we changed the data flow to reduce the burden of accessing SRAM and optimized the data arrangement. Referring to previous works [ 20 , 21 ], both used the number of spreading fingers for recognition, which is not accurate for gesture recognition.…”

Section: Resultsmentioning

confidence: 99%

“…It will consume extensive resources in hardware implementation. In our previous work [ 9 , 10 ], an implementation of FPGA for a 3D hand gesture architecture system under a complicated environment is provided.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Architecture Design and VLSI Implementation of 3D Hand Gesture Recognition System

Tsai

2021

Sensors

Self Cite

View full text Add to dashboard Cite

With advancements in technology, more and more research is being focused on enhancing daily life quality and convenience. Along with the increase in the development of gesture control systems, many controllers, such as the keyboard, mouse, and other devices, have been replaced with remote control products, which are gradually becoming more intuitive for users. However, vision-based hand gesture recognition systems still have many problems to overcome. Most hand detection methods adopt a skin filter or motion filter for pre-processing. However, in a noisy environment, it is not easy to correctly extract interesting objects. In this paper, a VLSI design with dual-cameras has been proposed to construct a depth map with a stereo matching algorithm and recognize hand gestures. The proposed system adopts an adaptive depth filter to separate interesting foreground objects from the background. We also propose dynamic gesture recognition using depth and coordinate information. The system can perform static and dynamic gesture recognition. The ASIC design is implemented in TSMC 90 nm with about 47.3 K gate counts, and 27.8 mW of power consumption. The average accuracy of each gesture recognition is 83.98%.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Resultsmentioning

confidence: 99%

See 1 more Smart Citation

Architecture Design and VLSI Implementation of 3D Hand Gesture Recognition System

Tsai

2021

Sensors

Self Cite

View full text Add to dashboard Cite

show abstract

“…A static gesture means keeping a certain pose without movement, whereas a dynamic gesture means a motion such as a swipe or a tap over sensors performed by users. Computer vision techniques are widely used to recognize gestures and human activities [10]- [13]. However, cameras can lead to privacy concerns for users, especially in residential environments, which limits their applications.…”

Section: A Sensors and Pattern Recognitionmentioning

confidence: 99%

An FPGA-Based Upper-Limb Rehabilitation Device for Gesture Recognition and Motion Evaluation Using Multi-Task Recurrent Neural Networks

et al. 2022

View full text Add to dashboard Cite

Upper-Extremity motor impairment affects millions of Americans due to cerebrovascular incidents, spinal cord injuries, or brain trauma. Current therapy practices used to assist these individuals in regaining motor functionality often require extensive time at rehabilitation facilities with potentially prohibitive travel or financial costs. This work presents a mobile low-cost field programmable gate array (FPGA)-smart rehabilitation system that can be used in home environments. The prototype is a rehabilitation table instrumented with a capacitive sensor array (CSA) to track upper-extremity motions of the user through proximity or touch. In addition, inertial measurement units (IMUs) are placed on the affected upper limb and combined with the CSA data with our sensor fusion signal processing architecture. Motions are classified and evaluated using multi-task convolutional recurrent neural networks with three additional motion quality output classes to personalize recognition based on the particular motor skills of each patient. The prototype achieves above 99% accuracy with 32-bit fixed-point format implementation for recognizing dynamic motions and identifying unnatural characteristics (i.e., tremor or limited flexion and extension) in upper limb motions based on sensor values. The convolutional recurrent neural network (C-RNN) fusion classification network is implemented on a 200 MHz Zynq ZCU104 FPGA using an HLS-based design optimized with pipelining and parallelism techniques and achieves 5.4x speedup compared to ARM ® Cortex-A53 implementation running at an operating frequency of 1.3 GHz. The prototype is also demonstrated to perform the machine learning classification in real-time.

show abstract

“…Stereo camera-based HGR systems [25]- [27] leverage the concept of human binocular vision to detect and interpret human hand movements. Using two synchronized cameras placed at a known distance apart, the system captures two slightly different images of a scene.…”

Section: Depth Camera-based Hgr Systemsmentioning

confidence: 99%

Design of low-power gesture recognition system and self-organizing map processors

View full text Add to dashboard Cite

Implementation of 3D Hand Gesture Recognition System using FPGA

Cited by 4 publications

References 3 publications

Architecture Design and VLSI Implementation of 3D Hand Gesture Recognition System

Architecture Design and VLSI Implementation of 3D Hand Gesture Recognition System

An FPGA-Based Upper-Limb Rehabilitation Device for Gesture Recognition and Motion Evaluation Using Multi-Task Recurrent Neural Networks

Design of low-power gesture recognition system and self-organizing map processors

Contact Info

Product

Resources

About