[360Labs.ai]
{ 360Labs }Case Study
{ Hardware and IoT }· Live

PanelVision

Automated visual quality inspection system for LG appliance control panels using computer vision to detect manufacturing defects by comparing production units against a reference template in real-time.

PanelVision
// overview

Camera-based visual inspection system for LG panels. SIFT and ORB feature detection for keypoint matching, homography-based alignment, SSIM differencing to quantify structural differences. Zero training data needed.

// specs

Specifications

CameraBasler ace 2 industrial GigE camera (5MP)
ComputeNVIDIA Jetson Xavier NX edge GPU
MethodSIFT/ORB keypoints + SSIM differencing
InnovationZero training data needed, template-comparison approach
// features

Features

  • 01Basler ace 2 industrial GigE camera with C-mount lens
  • 02Diffused LED ring light and backlight panel for uniform illumination
  • 03NVIDIA Jetson Xavier NX for on-site GPU inference with sub-second latency
  • 04Laser-based photoelectric trigger sensors synced to conveyor PLC
  • 05Pneumatic reject diverter actuator for flagged units
// tech stack

Tech Stack

PythonOpenCVNumPyscikit-imageSIFTORBSSIMEdge Computing

Interested
in PanelVision?

Let's discuss how this can work for you.

Get in Touch →
// end of case studyPanelVision
Contact