R&D Solutions
Software solutions in the field of computer graphics and Digital projects that allow you to implement tasks that are impossible at first glance.

Below you can see our cases.



Task: need to produce 750 standard clips in 10 days.
Organization of mass production of video cards of WorldSkills championship participants.

Need 3 types of clips with different design, design layouts provided by the customer. Each video includes personal information of the participant, a photo and several pictograms. All participant data was presented in the form of google tables. The peculiarity was that the data was updated continuously until the end of the project. In this connection, the task arose to track these updates. The data was textual information and a link to a random resource on which the participant's photo was located. In the data, there were anomalies that needed to be determined and edited. The photographs were of different quality, size with different framing and color characteristics, they had to be reduced to the same size, color profile and one type of framing.

The following Pipeline was developed in 1 day:

Given all this, it is extremely difficult and not efficient to complete this project on time manually, so it was decided to develop a fully automated content production pipeline
1
Data processing
Data from customer tables was automatically converted to machine-readable format, the converted data was then stored in our tables. Detection of changes in customer tables was implemented through the Google Scripts API trigger system.
2
IMG_Parse
Using a specially developed program, a parser written in python, photographs of participants were automatically extracted from the prepared data, which were then stored on local storage.
3
IMG_Processing
Using python and the OpenCV machine vision library, the position of the face was detected in the photos, according to the data received, the photo was cropped with the necessary aspect ratio. Using the Component Stretching algorithm in the image, the correct color balance was automatically restored. Images were scaled to the required size.
4
Automation rendering
Based on all the prepared data using the AfterEffects plug-in, DataClay Templater, the prepared scene was multiplied by many instances with the necessary changes within the project. Prepared instances were sent to batch-rendering.
5
Upload
Ready-made membership cards using a python script were uploaded to directories on the cloud to transfer content to the customer.
ORGANIZATION OF MASS PRODUCTION OF VIDEO CARDS OF WORLDSKILLS CHAMPIONSHIP PARTICIPANTS.
To contact us, please click on the telegram icon or email us at Petrukhin87@gmail.com

© 2021Cape Pro. All rights reserved. No part of this site may be reproduced without permission.