Image recognition in Python

Turning ideas into digital products

Gathering data via existing CCTV infrastructure. Object detection and pattern recognition in video streams for market research purposes. We’ve used machine learning techniques combined with power of OpenCV library in order to help companies better understand their target customers.

Goals

  1. Research a possibility to utilize image recognition in Market Research
  2. Preparation and tweaking of image recognition algorithms
  3. High accuracy of object detection

Process

  1. Started with a simple tool in OpenCV to detect people
  2. The tool was then updated to not use OpenCV’s standard detection algorithm, but based on an SVM (Support Vector Machine) machine learning algorithm to widen the range of objects that are being detected
  3. Then data was securely stored and explored in the database

Benefits

  1. Simplified and automated process of gathering data
  2. Completely new source of data about target customers
  3. Better understanding of customer profiles

Technology

Technology stack in this project includes:

Python

OpenCV

Git

SQL Server

We make digital products happen. Now.

GIVE ME A CALL

Hello, I am ready to answer all questions and talk about your ideas.

LEAVE US A CONTACT

We answer within one hour during working hours between 09:00 and 17:00 CET (currently 01:33 CET)