Volume 10, Issue 4, April – 2025 International Journal of Innovative Science and Research Technology

ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/25apr992

Hand Gesture Video Navigation System

B. Sri Ramya1; Ch. Bhargavi2; P. Dhanumjaya3; P. N. J. S. Siri4;
K. Vyshnavi5; G. Kavya6; Y. Pujitha 7
1
Assistant Professor
1,2,3,4,5,6,7
Department of Computer Science Engineering, Sri Vasavi Engineering College, Tadepalligudem,
Andhra Pradesh, India

Publication Date: 2025/04/30

Abstract: Hand gesture video navigation system control provides a seamless and touch-free way to interact with multimedia
systems, enhancing user convenience and accessibility. This project presents an automated system that enables users to
control video playback using only hand movements, eliminating the need for physical remotes or touch screens. The system
recognizes specific gestures to perform essential functions such as playing and pausing videos, fast-forwarding and
rewinding, increasing and decreasing volume, and zooming in and out. A camera captures hand movements in real time, and
the system processes these gestures to execute corresponding video control commands. This intuitive approach enhances user
experience, making video navigation more efficient and responsive. This project contributes to the advancement of touchless
human- computer interaction, making video control more accessible and user-friendly across various applications. Here
MediaPipe and OpenCV are playing a key role in the development of Hand gesture video navigation system.

Key words: OpenCv, MediaPipe, Human ComputerInteraction (HCI), Gesture Recognition, Video playback

How to Cite: B. Sri Ramya Asst.Professor; Ch. Bhargavi; P. Dhanumjaya; P. N. J. S. Siri; K. Vyshnavi; G. Kavya; Y. Pujitha (2025), Hand Gesture
Video Navigation System. International Journal of Innovative Science and Research Technology,
10(4), 1781-1788. https://doi.org/10.38124/ijisrt/25apr992

I. INTRODUCTION Improving accessibility for people with disabilities or

mobility impairments is one of the main driving forces behind
With the rapid growth of human-computer interaction, this project. For those who might find it difficult to use
touchless technologies have become an innovative method to traditional input devices, this system offers an inclusive
improve convenience and accessibility for users. One strategy alternative by doing away with the need for physical remote
is the use of a hand gesture-based control system. This controls or touch-based interfaces. Additionally, touchless
approach allows users to navigate some aspects of their control mechanisms aid in limiting the spread of
digital interfaces naturally and seamlessly. This project contaminants and germs in settings like hospitals, public
outlines the design of a Hand Gesture-Based Video areas, and industrial settings where hygiene is a top concern.
Navigation System developed to provide users with a touch-
free experience navigating multimedia content. This system Beyond entertainment, gesture-based video navigation
works in place of a standard input system (e.g., remotes, finds use in industrial automation, smart homes, assistive
keyboards, touchscreens) to provide a new way of interacting technology, and healthcare. This project contributes
with a digital interface when touch is a nuisance or limitation significantly to the development of intelligent user interfaces
for user interaction. and ubiquitous computing, which in turn promotes the
creation of novel hands-free control mechanisms for a variety
In order to interpret particular gestures, the suggested of applications as industries embrace touchless interaction
system uses a camera to record hand movements in real time. more and more.
These movements are then processed using computer vision
and machine learning algorithms. Essential video control By offering a natural, effective, and intuitive method of
features like play, pause, fast-forward, rewind, volume controlling multimedia with hand gestures, this research
control, and zooming in and out are mapped to these ultimately aims to close the gap between human intuition and
identified gestures. The system provides a seamless and digital interaction. This system has the potential to
effective multimedia control experience by guaranteeing high revolutionize how users interact with digital content by doing
responsiveness and accuracy in gesture recognition. away with the need for traditional input devices and providing
a technologically advanced, easily accessible, and convenient
substitute for traditional navigation techniques.

IJISRT25APR992 www.ijisrt.com 1781

Volume 10, Issue 4, April – 2025 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/25apr992
II. LITERATURE SURVEY and recognizes gestures with 95.2% accuracy [6]. Siddharth
Swarup Rautaray suggested a research paper on A Vision
The most important phase of software development is based Hand Gesture Interface for Controlling VLC Media
the literature review. This will provide an overview of some Player in 2010 using Pyramid Lucas-Kanade Optical Flow,
of the early research that many authors have done on this PCA, K-Mean, KNN Algorithms. He came to the conclusion
pertinent topic. We will also consider some important papers that using hand gestures, this application presents a clever,
and keep refining our work. touch-free interface for VLC control. By allowing users to
control the media player from a distance without a keyboard
Anklesh G suggested a research paper on Hand Gesture or mouse, it improves accessibility and convenience [7].
Recognition for Video Player in 2024 using OpenCV and Serkan Genç suggested a research paper on HandVR: a hand-
flutter. He came to the conclusion that this research makes it gesture-based interface to a video retrieval system in 2014
possible for video players to recognize hand gestures with using C++ using the OpenCV library. He came to the
great precision, guaranteeing effective operation. It improves conclusion that We investigated how well a hand-based
user involvement by providing a natural and easy way to interface could formulate queries for video retrieval systems.
control media playback with the usage of OpenCV [1]. Sakshi Some queries are difficult for traditional mouse-based
Shinde suggested a research paper on Gesture Based Media interfaces to handle because of their restricted flexibility in
Player Controller in 2022 using OpenCV. She came to the interaction. In order to improve engagement, we suggested a
conclusion that this program improves accessibility and hand-gesture-based interface that capitalizes on the innate
convenience by providing a touch-free interface for hand dexterity of human hands [8]. Ahmad Puad Ismail suggested
gesture-based VLC control. For smooth media management, a research paper on Hand gesture recognition on python and
it enables users—including those with disabilities—to OpenCV in 2020 using OpenCV and Haar cascade method.
personalize movements [2]. Shruti tibhe suggested a research He came to the conclusion that In summary, the project
paper on Media Controlling Using Hand Gestures in 2023 effectively used Python and OpenCV to recognize hand
using OpenCV and MediaPipe. She came to the conclusion gestures by utilizing hand segmentation and detection
that One more user-friendly and effective method of methods with the Haar-cascade classifier. In line with the
managing media players is by using hand gestures. Using project's aims, it accomplished two major goals: (1) building
Python and OpenCV, the system suggested in this project can a comprehensive system for computer vision-based hand
recognize hand gestures in real time and utilize them to gesture detection, recognition, and interpretation, and (2)
operate a range of media devices. The system may find use in producing a system that can identify numbers and sign
domains including gaming, public exhibitions, and home language movements[9]. Yuting Meng suggested a research
entertainment [3]. Manjunath R Kounte suggested a research paper on Real-Time Hand Gesture Monitoring Model Based
paper on Video Based Hand Gesture Detection System Using on MediaPipe’s Registerable System in OpenCV,MediaPipe
Machine Learning in 2022 using 2D CNN and Temporal Shift and FingerNet. He came to the conclusion that This study
Module (TSM). He came to the conclusion that in this introduces the Registerable Gesture Recognition Dataset
research, a hardware-efficient dynamic gesture detection (RGDS), which consists of 1,600 photos and 32 gesture types.
system for smart device control utilizing a Temporal Shift To improve feature extraction, a canonical transformation-
Model and CNN is proposed. By facilitating communication based normalizing technique was presented. The suggested
through natural hand gestures, it improves accessibility for FingerComb block sped up model convergence and enhanced
individuals with disabilities. The model is intended for feature robustness. Furthermore, RGDS was used to test and
gesture-based PowerPoint slide control and can identify construct an improved ResNet-based FingerNet model, which
sixteen different gestures [4]. Rishabh Agrawal suggested a showed excellent gesture recognition accuracy. For realistic
research paper on Real Time Hand Gesture Recognition for gesture-based interactions, these developments provide a
Human Computer Interaction in 2016 using OpenCV API in workable approach[10].
C++ language. He came to the conclusion that the suggested
technique for real-time hand gesture identification yields III. EXISTING SYSTEM
excellent outcomes with great precision and accuracy, and it
may be applied in a practical setting for engagement with a To enable user-computer contact, the majority of the
computer. It can precisely identify the fingertips, is far easier existing computer interaction systems rely on new technology
to use than a mouse, and has a lot of potential for expansion and physical input devices. Even while these gadgets and user
into other HCI applications [5]. Saransh Sharma suggested a interfaces have changed a lot over time, they still have built-
research paper on A Static Hand Gesture and Face in drawbacks that affect accessibility, usability, and
Recognition System for Blind People in 2019 using Haar efficiency.
cascade method, Linear Binary Pattern (LBP), Convex hull
and convex defects algorithms. He came to the conclusion  Wireless and Wired Mice
that with the use of LBPH recognizers and Haar cascade Because of their accuracy and ergonomic design, mice
classifiers for real-time face recognition, this system continue to be one of the most popular input devices.
functions as a virtual assistant for the blind and visually Conventional wired mice use a USB connection, which
impaired. The Convex Hull and Convex Defects approach is guarantees continuous use without battery issues. However,
used to identify hand gestures, and the YCbCr color scheme obstacles include limited mobility and cable clutter. Although
guarantees precise skin identification in a range of lighting wireless mouse with 2.4 GHz or Bluetooth communication
conditions. The system identifies faces with 92% accuracy provide greater flexibility, they also come with drawbacks

IJISRT25APR992 www.ijisrt.com 1782

Volume 10, Issue 4, April – 2025 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/25apr992
such possible latency, battery dependence, and network  Thumbs-Up Gesture:
interference. Furthermore, extended mouse use can lead to By using a thumbs-up, you can tell the system to start
repetitive strain injuries (RSI), which is why ergonomic playing the movie without using a remote control or tapping
designs have been developed. the screen.

 Trackpads and Touchpads  Fist Gesture:

Commonly featured in laptops, integrated touchpads To end a video or shut down a media player, clench your
enable scrolling, multi-touch gestures, and cursor movement fist.
without the need for an external device. To improve the user
experience, some sophisticated touchpads offer haptic  Volume Control:
feedback and force touch. Touchpads, on the other hand, may You can map hand actions, like raising or lowering a
not be as effective for extended work, lack the accuracy of a hand or rotating a fist, to change the volume levels.
dedicated mouse, and cause discomfort for those used to more
conventional pointing devices. Although they are frequently  Hand Tracking in Real Time with Computer Vision
restricted to particular environments, external trackpads
provide an experience comparable to built-in ones.  Live video input from a webcam or built-in camera is
processed by the system.
 Interfaces with touchscreens  It uses computer vision techniques based on deep
Touchscreens allow for direct device engagement, they learning to detect and track hand movements.
have completely changed computing. A common feature of  For real-time tracking and gesture classification, libraries
smartphones, tablets, and certain computers, they offer an like OpenCV, MediaPipe, and TensorFlow are used.
easy-to-use interface that supports tapping, pinching, and
swiping. Although touchscreens are incredibly portable and  Libraries for Computer Vision
user-friendly, they are less appropriate for applications that
demand a high level of precision, including complicated data
 OpenCV:
entry or graphic creation. In addition to displays being prone
An open-source package for tracking in real time,
to smudges and scratches, extended touchscreen use can wear
object detection, and image processing.
out fingers.
 MediaPipe:
 Voice Command
A framework created by Google that is ideal for
AI-powered assistants like Siri, Google Assistant,
gesture recognition and real-time hand tracking.
Alexa, and Cortana, voice-based communication has become
more and more common. These technologies enable users to
manage smart devices, dictate text, and carry out tasks. Voice
control is very helpful for accessibility since it allows people
with disabilities to use devices hands-free. However, because
voice recognition systems need internet access and data
processing, which raises security threats, they have
limitations in terms of speed, precision, and privacy.
Furthermore, voice instructions are not the best option for
intricate jobs like programming or design, and background
noise might affect accuracy.

IV. PROPOSED SYSTEM

Real-time hand gesture detection is used by a gesture-

based video navigation system to carry out different orders.
Among the main features are the following:

 Video Control Using Gesture Recognition

 Swipe Gesture:
Without pushing buttons, you may easily move through
material by swiping left or right to fast-forward or rewind the
movie.

 Hand Open Gesture:

Without having to look for a pause button, you can
easily interact with the video by pausing it with an open hand.

IJISRT25APR992 www.ijisrt.com 1783

Volume 10, Issue 4, April – 2025 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/25apr992
V. IMPLIMENTATION PROCESS DIAGRAM

Fig 1 Implementation

Figure 1 Implementation illustrates the step-by-step  Feature Extraction

process involved in hand gesture recognition for a media Finding important details in the hand image to
player. differentiate between various gestures depends heavily on
feature extraction. By identifying 21 important landmarks on
 Data Acquisition each hand, MediaPipe Hand Tracking provides real-time
Real-time hand detection and tracking using a webcam tracking with little computing overhead and allows for
is the only way the gesture-based video navigation system accurate finger position and orientation recognition.
works. Without requiring labeled training data, the system Furthermore, form and texture data are extracted using the
recognizes hand motions by dynamically processing live Histogram of Oriented Gradients (HOG) approach, which
video frames. In order to maximize performance, OpenCV strengthens gesture detection in a variety of backgrounds.
extracts frames at a predetermined rate from continuously Edge detection techniques like Sobel and Canny are used to
recorded video input from a camera. Advanced computer draw attention to hand boundaries in order to increase
vision methods like MediaPipe Hand Tracking are used for accuracy. This helps the system better identify different hand
hand detection, which enables the system to identify hands forms and movements.
quickly without consulting existing datasets. While Haar
Cascades provide pattern-based detection for real-time  Model Training and Gesture Recognition
recognition, MediaPipe offers exact tracking by recognizing To guarantee precise hand gesture identification, the
21 important landmarks on the hand. After the hand has been system makes use of a Convolutional Neural Network
identified, the features that have been retrieved are examined, (CNN). Preprocessed images are received by the input layer,
and gesture-based commands are mapped to operate the video which then transforms them into an appropriate processing
playback. format. Convolutional layers extract crucial spatial data from
the photos by applying many filters to identify edges,
 Preprocessing textures, and distinctive hand patterns. By reducing
The collected frames are preprocessed before being dimensionality while maintaining important features, a
entered into the machine learning model in order to improve pooling layer (Max Pooling) increases computational
accuracy. The first stage is RGB conversion, in which the efficiency. After being retrieved, the features are converted
Luminance Method is used to convert frames from OpenCV's into a one-dimensional vector by passing them through fully
default BGR format to RGB: linked layers. Lastly, the output layer determines the detected
Y=0.2989R+0.5870G+0.1140B. Better feature extraction is hand movement with high accuracy by assigning probabilities
ensured by this conversion, which also helps to modify to various motions.
contrast and visibility. After that, resizing is done to keep all
input data consistent by standardizing photos to a fixed  Video Navigation Control
resolution, like 128 by 128 pixels. Lastly, undesired Once gestures are recognized, they are mapped to
distortions are removed using noise reduction techniques like specific video playback controls.
Gaussian blur or median filtering, which improves hand
detection accuracy and system performance in general.  Five→ Play/Pause
 One/Two→ Forward/Rewind
 Three/Four→ Volume Increase/Decrease

IJISRT25APR992 www.ijisrt.com 1784

Volume 10, Issue 4, April – 2025 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/25apr992
 Wrist → Zoom in confirming precise gesture detection from various
 Five fingers move backward → Zoom out viewpoints. Testing involves several users with different hand
sizes, skin tones, and orientations. Generalization is enhanced
 Testing and Optimization by data augmentation methods including flipping, rotating,
Real-world testing is done on the system to improve its and introducing noise. The hand tracking parameters are
performance and accuracy. To guarantee flexibility, lighting adjusted for increased efficiency because deep learning is not
conditions are assessed in a range of indoor and outdoor utilized. Algorithm improvement improves real-time
environments. For accurate hand detection, backdrop execution speed and accuracy. Smooth operation on common
variations are evaluated with both simple and complicated hardware is ensured by computational load reduction. To
backgrounds. Variations in distance and angle aid in increase accuracy and resilience, the system is iteratively
improved depending on test results.

VI. EXPERIMENTAL RESULTS

Fig 2 Forward

In the above image (Figure2), the recognized gesture successfully triggers the forward action, skipping the video by 5
seconds.

Fig 3 Backward

In the above image (Figure 3), the recognized gesture successfully triggers the backward action, rewinds the video by 5
seconds.

IJISRT25APR992 www.ijisrt.com 1785

Volume 10, Issue 4, April – 2025 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/25apr992

Fig 4 Volume Increase

The system accurately detects the hand gestures and increases the video volume accordingly. This enhances user interaction
by providing seamless volume control through gestures (Figure 4).

Fig 5 Volume Decrease

The system recognizes the hand gesture and decreases the video volume accordingly. This enables smooth and efficient volume
control using intuitive hand movements (Figure 5).

Fig 6 Play/Pause

IJISRT25APR992 www.ijisrt.com 1786

Volume 10, Issue 4, April – 2025 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/25apr992

The system detects the hand gesture and pauses the video instantly. This provides a hands-free and seamless way to control
video playback (Figure 6).

Fig 7 Zoom in

The system recognizes the fist gesture and triggers the zoom-in function. This reduces the content size, allowing a wider view
of the screen (Figure 7).

Fig 8 Zoom out

The system detects the five-finger backward movement infotainment systems can all benefit from the technology. For
gesture and activates the zoom-out function. This enhances a seamless user experience, it guarantees real-time response
the viewing experience by enlarging the content for better and avoids unwanted behaviors. Digital media control is
visibility (Figure 8). about to be redefined by advances in AI and gesture detection.
Gesture-based navigation will be a common solution as
VII. CONCLUSION contactless technology advances. Digital interactions are now
more intelligent, effective, and futuristic thanks to this
The hand gesture-based video navigation system offers invention.
a simple, touch-free method of controlling media playing,
revolutionizing human-computer interaction. It tracks hand REFERENCES
landmarks and recognizes gestures using OpenCV and
MediaPipe, and PyAutoGUI maps these movements to media [1] Anklesh G, Akash V, Prithivi Sakthi B ,
instructions. Simple hand gestures can be used by users to Kanthimathi.M – “Hand Gesture Recognition for
navigate, play, pause, and adjust volume. This method Video Player” in 2024.
improves accessibility and provides those with mobility [2] Sakshi Shinde, Sarthak Mushrif, Aditya Pardeshi,
disabilities with an inclusive solution. By decreasing physical Dhairyasheel Jagtap , Guide: Prof. Vandana Rupnar –
touch in public areas, it also acts as a hygienic substitute. “Gesture Based Media Player Controller” in 2022.
Virtual reality, gaming, public displays, smart homes, and car

IJISRT25APR992 www.ijisrt.com 1787

Volume 10, Issue 4, April – 2025 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/25apr992
[3] Shruti Tibhe, Ashwini Joshi, Aishwarya Warulkar,
Aishwarya Sonawane , Miss. T.U.Ahirrao – “Media
Controlling Using Hand Gestures” in 2023.
[4] Manjunath R Kounte, E Niveditha , A Sai Sudeshna ,
Kalaigar Afrose – “Video Based Hand Gesture
Detection System Using Machine Learning” in 2020.
[5] Rishabh Agrawal, Nikita Gupta – “Real Time Hand
Gesture Recognition for Human Computer
Interaction” in 2016.
[6] Saransh Sharma, Samyak Jain, Khushboo – “A Static
Hand Gesture and Face Recognition System For Blind
People” in 2019.
[7] Siddharth Swarup Rautaray, Anupam Agrawal – “ A
Vision based Hand Gesture Interface for Controlling
VLC Media Player” in 2010.
[8] Serkan Genç, Muhammet Ba¸stan, Ugur Gudukbay
,Volkan Atalay, Ozgur Ulusoy – “HandVR: a hand-
gesture-based interface to a video retrieval system” in
2014.
[9] Ahmad Puad Ismail, Farah Athirah Abd Aziz, Nazirah
Mohamat Kasim, Kamarulazhar Daud – “Hand gesture
recognition on python and opencv ” in 2020.
[10] Yuting Meng , Haibo Jiang, Nengquan Duan, Haijun
Wen – “Real-Time Hand Gesture Monitoring Model
Based on MediaPipe’s Registerable System” in 2024.

IJISRT25APR992 www.ijisrt.com 1788