Skip to main content
    Scalexa — Senior Engineering & AI Solutions
    Computer Vision

    Track and Blur Objects

    Project Overview

    We built a production-grade video analysis platform for Investigation VE, a specialist provider serving law enforcement and investigative agencies. The system enables rapid detection, tracking, and redaction of sensitive elements in video footage—turning hours of manual work into minutes of automated processing.

    Track and Blur Objects project screenshot

    The Challenge: Manual Video Processing at Scale

    Investigation VE needed to offer their law enforcement clients a faster, more reliable way to process video evidence. Existing workflows required analysts to manually scrub through hours of footage, frame by frame, to identify and redact faces, license plates, and other sensitive information. This was expensive, error-prone, and couldn't scale with demand.

    What We Built

    We delivered a complete video analysis platform powered by YOLOv10, with purpose-built tooling for investigative workflows.

    • Automated detection of faces, people, weapons (knives, guns), and license plates
    • Frame-accurate tracking that follows subjects across the full video timeline
    • One-click blur or erase functionality for privacy redaction
    • Cloud-based processing using AWS Batch for parallel video handling
    • Cross-platform Electron desktop application for secure local use
    • WebSocket-based real-time collaboration for team review sessions
    • Multi-format export (MP4, AVI, MOV) for compatibility with evidence systems

    Technical Architecture

    The platform combines edge performance with cloud scalability.

    • YOLOv10 inference pipeline optimised for video streams
    • AWS Batch for distributed processing of large video files
    • S3-based asset storage with presigned URLs for secure access
    • Electron frontend with native file system integration
    • WebSocket server for real-time sync between collaborators
    • Usage-based billing engine tied to video resolution and duration

    The Impact

    • Processing time reduced from hours to minutes per video
    • Eliminated manual frame-by-frame scrubbing for redaction tasks
    • Enabled Investigation VE to scale capacity without adding headcount
    • Delivered a SaaS-ready platform with usage-based pricing model

    Why It Matters

    This engagement gave Investigation VE a competitive edge in the law enforcement technology space. By automating the most time-consuming parts of video evidence processing, their clients can focus on investigations rather than manual editing. The platform continues to process video for agencies across the sector.

    "My project manager collaborated closely with Scalexa and reported that the experience was pleasant and productive."

    AS
    Andrew Simon
    Founder, Klever Thoughts Inc.
    Get Started

    Ready to Get Started?

    Book a free 30-minute discovery session with our senior engineers to identify quick wins and show you what's possible.

    View Our Work