Skip to content
Revolutionizing business using computer Vision

Revolutionizing business using computer Vision

Garvita Jain, Sumant Kulkarni

Computer Vision is a field of Artificial Intelligence which enables computers to interpret and analyze the visual world with better efficacy. It has gained immense popularity in past few years in dynamic industries such as retail, insurance and manufacturing. These industries are leveraging machine vision to enhance their customer experience, reduce time and efforts and achieve better quality assurance.


It is well acknowledged that the retail industry is at the forefront in leveraging computer vision. This would help improve customer experience and provide relevant data and insights to retailers. With increasing popularity of ecommerce, businesses are evolving to offer customer delight by leveraging computer vision for personalized and streamlined in-store shopping experience. Computer vision allows retailers to speed up business operations like shelf management, payments and data collection. Let’s talk about some integral computer vision solutions that Zensar has built for our retail customers.

Facial Recognition System

Every retail store has cameras for security reasons. These cameras can be used to recognize faces and identify frequent customers and new customers. This identification can help retailers to give discounts to increase brand loyalty and to attract new customers. The simplest way of attracting new customers is by providing them most suitable recommendations basis their purchase history. To put this to use, Zensar has built an in-house solution using computer vision algorithms which registers faces of people in the system. When the same person visits the store, he/she gets recognized by the system.


A deeper analysis of footfall, customer demographics and user satisfaction level can help retailers identify their most popular products which allow them to re-design their store layout. As an attempt to promote the sales of the less consumed products we’ve built a platform called ZenVAS (visual analytics system) which identifies age, gender and emotion of the customer. It tracks the movements of people and in turn reveals useful insights such as most popular zone (product), most popular path followed in the store by people and most popular time of day for purchase.

comp vision pic 1


comp vision pic 2

Reverse Image Search

Customers often come across something that they want to buy, but somehow; do not have relevant information about it. Object recognition technology can be used to recognize such products and provide contextual information about it. It can also direct the user to same/similar product. ‘Try image search’ option has got wide acceptance by customers in many popular ecommerce sites.  Zensar has advanced expertise in ‘reverse image search’ feature for clothes recognition that can be used by e-retailers and can be extended to cover other object types as well.


Claim Processing in Insurance is a time-consuming process and relies a lot on human intervention. After a claim has been filed, a human adjuster visits the workshop (in case of asset damage) or the place where the damage occurred (in case of home insurance) to inspect the damage, validate claim and coverage, evaluate the claim amount and approve payment followed by the finance department initiating payment.

Computer vision can play a vital role in eliminating the roadblocks in faster processing of claims by doing automatic damage detection and assessment.

ZenLabs, our premier innovation hub, in collaboration with Cynosure, a Zensar company has built two First Notice of Loss (FNOL) solutions. You can read below to learn more about these two solutions.

Car Damage Assessment

This in-house solution fastens the claim processing for car damage by doing auto-detection of damaged parts and auto-assessment of severity of damage to estimate the claim amount. The user can login using his/her credentials on the app. The details of the user such as name, policy number and vehicle number get populated from guidewire. The user can then use photo claim option to take pictures of damaged cars. The AI engine analyzes those images, identifies the damaged parts of car and assesses the severity of damage. Based on this assessment, claim amount is evaluated. If the user is satisfied with the estimates, he/she can submit the claim for processing to guidewire. 

Roof Damage Assessment

This solution is a part of the home insurance claim processing and identifies the part of the roof which is damaged due to hailstorms or any other natural calamity. The pictures are taken using drones and assessed using computer vision algorithms. 


Quality assurance is the most expensive activity in production and manual inspections are carried out for the same. Computer vision makes it possible to spot minor defects that are not visible to the human eye. According to Forbes, AI can improve manufacturing defect detection rate by 90%. Surface Imperfection Detection is a quality assurance task which is mandatory to guarantee the quality of a manufactured item.

Steel Defect Identification

Defect Identification on steel sheets is one such step in reducing the manual efforts in quality check. Surface defects on steel sheets are not identifiable by human eyes and requires the use of high frequency cameras to detect the same. We have built a solution to localize and classify those defects in four categories using computer vision algorithms.


Artificial Intelligence is disrupting business and society in a pivotal way. Computer vision is enabling a multitude of industries like retail, insurance, manufacturing, etc. to achieve enhanced customer delight and satisfaction. Our ‘Living AI’ philosophy, compels us to empower businesses to provide better services to their customers.

Data Engineering & Analytics

Also read: