profile-pic

Akshara P K

I’m a Computer Vision & Image Processing Engineer with 9+ years of experience building real-world machine vision solutions across industrial automation, medical imaging, and AI-driven inspection systems.


Currently, I work as a Lead Engineer – Image Processing at Dover India, where I contribute to a Vehicle Hail Scan System focused on detecting dents on vehicle surfaces. My work spans 3D reconstruction, 3D-to-orthographic conversion, panel identification, dent detection using YOLO, dockerizing training/inference pipelines, and drift analysis.


Across my career, I’ve built and delivered end-to-end solutions using C++ and OpenCV, along with Python-based deep learning pipelines (TensorFlow/Keras). I’ve also worked closely with cross-functional teams including optical engineers, research teams, and customers across Japan and Taiwan.


I enjoy solving complex vision problems, building scalable algorithm pipelines, and mentoring engineers who are entering the field of computer vision.

  • Role

    Lead Engineer - C++

  • Years of Experience

    11.67 years

Skillsets

  • machine vision
  • Computer Vision
  • Docker
  • Git
  • Image Processing
  • Image transformations
  • Jira
  • Linux
  • Machine Learning
  • clustering
  • Medical Imaging
  • object detection
  • Object Tracking
  • R
  • Retinanet
  • Segmentation
  • SVN
  • Mesh construction
  • TensorFlow
  • C++
  • Python
  • MATLAB
  • OpenCV
  • Open3D
  • Scikit-image
  • SciPy
  • NumPy
  • Windows
  • Keras
  • Pattern recognition
  • Faster-rcnn
  • Yolo
  • Deep Learning
  • Python
  • 3D reconstruction
  • C

Professional Summary

11.67Years
  • Jun, 2023 - Present2 yr 11 months

    Senior Software Engineer - Medical Imaging

    Advitech Technologies
  • May, 2021 - May, 20232 yr

    Lead Engineer - Image Processing

    Dover India
  • Oct, 2018 - May, 20212 yr 7 months

    Senior Engineer-Computer Vision

  • Jun, 2014 - Apr, 2015 10 months

    Technical Assistant

    Indian Institute of Information Technology and Management- Kerala
  • Apr, 2015 - Sep, 20183 yr 5 months

    Senior Software Engineer

Work History

11.67Years

Senior Software Engineer - Medical Imaging

Advitech Technologies
Jun, 2023 - Present2 yr 11 months

Lead Engineer - Image Processing

Dover India
May, 2021 - May, 20232 yr
    Works for Vehicle Hail Scan System intended to detect dents caused by hails on vehicle surfaces. Responsible for implementing image processing algorithms for 3D reconstruction, 3D to orthographic conversions, panel identification, dent detection using YOLO, dockerizing training and inference pipelines, drift analysis. Extensively used C++ and OpenCV. Uses python for document analysis, PaddleOCR for table recognition and trained detection models for OCR text detection.

Senior Engineer-Computer Vision

Oct, 2018 - May, 20212 yr 7 months

Senior Software Engineer

Apr, 2015 - Sep, 20183 yr 5 months
    Developed computer vision algorithms for human-machine interaction. Worked with imagineers and optical engineers to provide end-to-end solutions for automation in semiconductor manufacturing, robotics, medical and food industry. Collaborated with semiconductor manufacturing corporations internationally.

Technical Assistant

Indian Institute of Information Technology and Management- Kerala
Jun, 2014 - Apr, 2015 10 months
    Project: Application of Reverse Diffusion for vessel enhancement in Multi-channel Phase Contrast Angiography. Restoration of vessel structures in multi-channel Phase contrast Angiography with Non-Locally Stabilized Reverse Diffusion.

Achievements

  • First Rank in Bachelor of Computer Applications from Mahatma Gandhi University
  • Second Rank in Master of Computer Applications from Mahatma Gandhi University
  • Second Rank in Master of Philosophy in Computer Science from Cochin University of Science and Technology Received GES Spot Champ on the basis of customer feedback proving excellent work at end customer site in completing project successfully Received Kimball Spot Champ for the excellence in guiding vision team which lead to success of critical project

Major Projects

10Projects

Alignment Correction Solutions

    Developed generic algorithms to align the device under test for further processing. Implemented five different algorithms for alignment. Implemented in C++ using OpenCV.

Pattern Recognition Solutions

    Developed generic algorithms to identify a target in an image with reference to a template pattern. Implemented five conventional algorithms for pattern matching. Implemented in C++ using OpenCV.

Cosmetic Inspection on Optical Stabilizer

    Designed and implemented an algorithm to detect dust, pit, discoloration, and scratches on metal surfaces of optical stabilizers for mobile cameras. Conventional algorithm exploiting gray scale deviations. Implemented in C++ using OpenCV.

Foreign Object Presence Detection

    Developed a package with acquisition options and a generic algorithm to detect and locate foreign objects within a device under test. Includes a teaching environment for user-defined ROIs and exclusion areas. Implemented in C++ using OpenCV.

Image Classification

    Generic deep learning solution for classification problems allowing users to create and train their own CNN models. Features include network design, training progress, data augmentation, testing, and confusion matrix generation. Implemented in Python using TensorFlow.

Medical Packet Inspection System

    Machine learning solution for defect detection in transparent medical packets using a fine-tuned FasterRCNN architecture. Implemented in Python using TensorFlow.

Metrology Solutions for Wafer Images

    Developed algorithms to measure surface roughness and etch-depth of wafers for Automated AFM tool. Includes post-production inspection algorithms. Implemented in C++ using OpenCV.

Virtual Guided Assembly

    Developed a vision-guided assistance solution for manual PCB assembly using pattern matching techniques. Implemented in C++ using OpenCV.

Bank Details Extraction

    Extracts text from images and filters required information using PyTesseract, Google Cloud API, and spaCy for Named Entity Recognition. Implemented in Python; interface used is Google Colab.

Unsupervised Objects Clustering

    Categorizes objects into classes using preprocessing, feature extraction (SIFT, HOG, LBP), and clustering (KMeans, distance measures). Implemented in C++; interface used is Google Colab.

Education

  • Master of Philosophy in Computer Science

    Indian Institute of Information Technology and Management, Kerala (2014)

  • Master of Computer Applications

    Mahatma Gandhi University, Kerala (2012)

  • Bachelor of Computer Applications

    Mahatma Gandhi University, Kerala (2009)

Certifications

  • SPIE Journal of Medical Imaging 3(1), 014001 (22 January 2016)

Interests

  • Machine Vision