Minerva
Theia Logo
Image
Analytics
Analytics
Theia empowers businesses to gain insight on images and video data through computer vision techniques. It is capable to acquire images/videos data from various sources, process it using deep learning models and make inferences. It can be used across industries to enhance the consumer experience, reduce costs and control quality.
Theia automates tasks like image classification, image segmentation, object detection, and feature matching that human visual system can do.
Theia empowers with enhanced text extraction using computer vision capabilities and visual data processing. It helps to classify content, extract printed and handwritten text. Recognise familiar areas like brands and landmarks and watermark content
Theia automates tasks like image classification, image segmentation, object detection, and feature matching that human visual system can do.
Theia empowers with enhanced text extraction using computer vision capabilities and visual data processing. It helps to classify content, extract printed and handwritten text. Recognise familiar areas like brands and landmarks and watermark content
Success stories of Theia
0.2 sec.
Time to confirm Deduplication check for 1000 images
87%
Accuracy achieved to detect images having watermarks
3x
Increased in product match rate for Competitive Intelligence
97%
Accuracy achieved in Image auditing covering 7 quality checks
Features of theia

Image Auditing
- Standard Image Size
- Blur Detection
- File Format Validation
- Black & White Color
- De-duplication
- Person in Image Detection
- Watermark on Image

Defect Detection
- ML-driven image classification to detect the quality issues.
- Reduces human effort and automatically classifies images as pass or fail.
- Fabric Defects Detection, Device faults detection through

Auto-cropping of images
- Identify the Subject in the image and crop the subject area
- Identify X & Y parameters based upon extremities of the subject
- Automatically Scale and crop as per required features

Image Meta Tagging
- Extract Meta tags from Image ranging from High level to detailed objects
- Identify detailed attributes of objects such as type, category, pattern, colour and fabric
- Identify emotions, such as, joy, sorrow

Background Removal
- To separate foreground image information.
- Remove background

Logo Detection
- Detect product logoswithin an image.
- Crop detected logo.
©2024 eClerx Services. All rights reserved.