iVideo Editor with Background Remover and Image Inpainting

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 09 Issue: 04 | Apr 2022 www.irjet.net p-ISSN: 2395-0072
© 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 2512
iVideo Editor with Background Remover and Image Inpainting
Pentaparthi Tapaswi1, Addenki Sri Pooja2, Mohammed Muskaan3, Harika Manthena4, Asst Prof.
K. Archana5
1,2,3,4Student, Dept. of IT., Gokaraju Rangaraju Institute of Engineering & Technology., Hyderabad, Telangana
5Asst. Professor, Dept. of IT., Gokaraju Rangaraju Institute of Engineering & Technology., Hyderabad, Telangana
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - Inpainting is a technique for reconstructing
damaged or defective areas by interpolating neighboring
pixels. To eliminate undesired components from an image,
utilize the Image Inpainting software. Photo to sketch,
backdrop removal, andvideoeditorareamongthecapabilities
of iVideo Editor, an online tool. We'll change the original RGB
values and assign them to grey, then invert and blurtheimage
to make it into a drawing. Remove.bg API is an image editing
service that removes the backdrop of people and product
photos. Trim, merge, preview, and render are someofthetools
available in a video editor.
Key Words: Image Processing, Background Remover,
Photo to Sketch, Fast Marching method, Navier Stokes
1. INTRODUCTION
1.1 Image Inpainting
Image inpainting is the technique of filling in missing data in
a specific region of the visual input.Theprocedure'spurpose
is to restore missing or damaged picture components to the
point that the inpainted region is imperceptible to the
untrained eye. The variety of applications ranging from the
reconstruction of incomplete blocks caused by packet loss
during wireless transmission tothe removal/introductionof
image objects such as stamped dates, logos, people, text and
special effects on the scene, to the reversal of impairments
such as cracks, scratches, and dirt in scanned photographs
and digitized artwork images,totheremoval/introductionof
image objects such as logos, stamped dates, text, people,and
special effects on the scene, to the removal/introduction of
image object. The Color Filter is another component of this
project. Color Filter refers to adding effects of RGB
(RED|GREEN|BLUE) solid layers on top of theoriginal layers
of image. We can see the significance of the color
combination in the real world. Color filtering is all about
modifying the RGB components.
1.2 iVideo Editor
ivideo editor is a web app with features like photo to sketch,
background remover and video editor. Web app is built
using flask which is rapid and low-level pythondevelopment
framework. You'll need to modify the image if you want to
make prints that go beyond plain records to capturethesoul
of individuals that stand alone as works of art. The creative
act of photographic printmaking revolves around image
editing. Our app removes the background from your photos
without lowering the quality. Photo to sketch convertsinput
image to grey image using cvtcolor function of OpenCV, next
invert image, blur image, invert blurred image, sketch, and
display sketch. Remove. Bg API isa photoeditingservicethat
eliminates the backdrop of people and product photos. The
output resolution is capped at 10 megapixels. For
authentication, API keys are required. The process of
modifying and arranging video clips is known as video
editing. Most films and television shows, commercials, and
video short stories, is structured and presented using video
editing. Video editing software has significantly
democratised video editing in recent years. Because video
editing can be difficult and time-consuming, many
technologies have been developed to help people with this
task. Our iVideo editor is one such application. Web based
video editor created using Moviepy and python. This
provides features of trim using time zones in a video,
preview video and file download to desired folder. Moviepy
is a video editing module that may be used to do operations
like as concatenations, cuts, processing, and effects. It
supports the most popular media formats, including gif, and
can read and write them.
2. Existing System
There are softwaressuchasAdobePhotoshopwhichprovide
image background removal option, image color effects.They
have a selection tool which allow us to select a particular
region and remove it. But these only work on high end
laptops (8GB minimum required). And this also requires a
skill and experience with cloning and effects in the
photoshop. Similarly, video editing softwares like Adobe
premiere pro require high end RAM and processor for
previewing and rendering the trimmed clip.
3. Proposed System
In our project, we combined photo editing (inpainting, color
filter, background remover, photo to sketch) features and
video editing (trim, add, render and preview) features
together with minimal requirements and less consumption
of RAM. This is entirely automatic and doesn’t require
human intervention. Artificial intelligence is used by the
Background Removertodeterminethephoto'ssubject,mask
the image, and eliminate the background. We use the API of
this AI tool.

4. Implementation and Methodology
4.1 Fast Marching
The Fast-Marching Algorithm starts at the region's edgeand
works its way in, filling everything in the edge first. A small
neighbourhood must be painted around the pixel on the
neighbourhood. The pixel in question is replaced by the
normalised weighted sum of all known pixels in the area.
The weights must be carefully chosen. Pixels close to the
point, nearer to the normal of the boundary, and on the
boundary contours are given more weight. When a pixel has
been inpainted, it moves to the next closest pixel using this
method. Because the Fast Marching technique ensures that
pixels nearby to known pixels are inpainted initially, it is
essentially a manual heuristic procedure.
4.2 Navier Stokes
Partially differential equations are used in this algorithm,
which is based on fluid dynamics. Heuristic is the
fundamental principle. It starts by exploring the outskirts of
known & unknown locations (as edges are meant to be
continuous). It persists isophones while matching gradient
vectors at the inpainting region's edge (lines linkingareas of
similar intensity, related to how contours connect points of
similar height). This is accomplished by employing fluid
dynamics methods. To decreasetheminimal variancewithin
this area, colour is filled in after they are received.
Fig -1: Architecture Diagram
4.3 Color Filters
The RGB color space is the most widely used; it is known as
an additive color space because the three-colortonesadd up
to provide the image color. The application of a color filter
alters the appearance of an image or a portionofanimageby
modifying the shades and colors of the pixels in some way.
Filters are used to enhance a photograph's brightness and
contrast, as well as to add a variety of textures, tones, and
special effects.
4.4 Photo to Sketch
For Photo to Sketch, use the cvtColor() function to convert
the image to grayscale. This function converts the colour
space of an input image. Then we can convert it to grayscale
to get a black and white image of our output. The function
bitwise not() performs per-elementbit-wiseinversionof the
input, converting brighter regions to lighter and lighter
regions to brighter, allowing us to easily find image
edges. The function GaussianBlur() is used to reduce image
noises and blurring smoothens the source image with the
specified Gaussian kernel. The dodgeV2() function divides
the grayscale image by the inverse of the blurred image,
leaving only the highlights and the most prominent edges.
4.5 Background Remover
Remove.bg API is a photo editing service that eliminates the
backdrop of people and product photos. The output
resolution is capped at 10 megapixels. For authentication,
API Keys are required.
4.6 Video Editor
Pillow is to images what MoviePy is to films. It has a variety
of features for typical operations such as importing, editing,
and exporting video files. It also allows you to place titles
into videos and rotate them 90 degrees (horizontal and
vertical flipping can also be implemented with this logic).
4.7 Flask
Flask is a microweb framework writteninPython.Itcould be
utilised to extend the functionality of an application as if it
were constructed into Flask. For example, extensions for
object relational mappers, form validation, upload handling,
several openauthentication protocols,andotherframework-
related tools are available. Applications like Pinterest and
LinkedIn use the Flask framework.
5. RESULTS
Fig -2: Image inpainting app

Fig -3: Masking unwanted section
Fig -4: Inpainted image using Fast Marching methodology
Fig -5: iVideo Editor Home page
Fig -6: Background removed image and sketched image
Fig -7: Video Editor
6. CONCLUSIONS AND FUTURE WORK
Remove.bg API internally uses machine learning to detect
edges of the subject, there by distinguishing it from the
background. So, the portrait picturewithoutbackground can
be obtained by replacing this layer with empty pixels. Many
filters are required to convert image into pencil sketch, but
our webapp converts it automatically using OpenCV. Video
editor is being developed using flask because it is a small-
scale application and MoviePy module was used to perform
editing. This can decrease the burden on RAM from opening
heavy softwares like adobe premiere pro for small
operations like Trim, preview, and rendering.Forimage size
compression, the existing tools reduces the size
compromising on quality. In future, we would like to add
Image size compression feature without losing quality.
REFERENCES
[1] Alexandru Telea, “An ImageInpaintingTechniqueBased
on the Fast-Marching Method”, Vol 9,
10.1080/10867651.2004.10487596, Journal of Graphics
Tools, Eindhoven University of Technology, 2004

[2] Bertalmio, M. & Bertozzi, Andrea & Sapiro, G.. (2001).
Navier-Stokes, fluid dynamics, and image and video
inpainting. Proc. IEEE Comput. Soc. Conf. Comput. Vis.
Pattern Recognit. 1. I-355.10.1109/CVPR.2001.990497,
2001.
[3] Converting Images to Pencil Sketch :
“https://www.askpython.com/python/examples/image
s-to-pencil-sketch”.
[4] API for background Removal :
“https://www.remove.bg/api”.
[5] Video Editor :
“https://www.geeksforgeeks.org/introduction-to-
moviepy”.

iVideo Editor with Background Remover and Image Inpainting

More Related Content

Similar to iVideo Editor with Background Remover and Image Inpainting (20)

More from IRJET Journal (20)

Recently uploaded (20)

iVideo Editor with Background Remover and Image Inpainting