Deployment
R2026bSince R2026b
The Visual Inspection Toolbox™ provides capabilities for deploying trained deep learning models to production environments and external inference frameworks.
Export models to ONNX — Export trained visual inspection models to the Open Neural Network Exchange (ONNX) format for integration with external inference frameworks. The Student-Teacher anomaly detector exports as an end-to-end model. The YOLOX object detector only exports the core detection network, with separate MATLAB® functions for preprocessing and postprocessing that handle resizing with a preserved aspect ratio, input formatting, threshold filtering, and non-maximum suppression. These functions support code generation for deploying complete detection systems on edge hardware.
Generate code — The toolbox supports deployment through CPU and GPU code generation. You can use MATLAB Coder™ to generate optimized C or C++ code for CPU-based deployment, or GPU Coder™ to generate CUDA® code for high-performance execution on NVIDIA® GPUs. These code generation tools enable you to package trained networks, along with preprocessing and postprocessing logic, into deployable libraries and executables. For more details, see MATLAB Coder (MATLAB Coder) and GPU Coder (GPU Coder).
Create standalone applications — You can use MATLAB Compiler™ to package visual inspection pipelines as standalone applications or deployable archives for sharing with end users who do not have MATLAB.
Deploy to field-programmable gate array (FPGA) — Deploy a YOLOX object detector to FPGA hardware for low-latency object detection.
These capabilities support a variety of deployment scenarios, including edge device inference, cloud-based prediction services, and integration with third-party frameworks, providing a pathway for validating deployment and ensuring that deployed models maintain the accuracy and behavior established during development. For more information, see Deploy Visual Inspection Code, Models, and Applications.
Functions
Topics
- Deploy Visual Inspection Code, Models, and Applications
Deploy visual inspection code, models, and applications to production environments using ONNX export, code generation, or standalone application packaging.
Featured Examples
Deploy ONNX YOLOX Object Detector on NVIDIA Jetson Orin Nano for PCB Defect Detection
Deploy an ONNX YOLOX object detector on an NVIDIA® Jetson™ Orin Nano board to detect and localize defects in a printed circuit board (PCB). Defects in PCBs can cause performance degradation or product failure, so it is important to detect PCB defects early in production. Identifying faulty boards early in the production process helps manufacturers maintain high product quality and prevent downstream issues.
Code Generation for Detect Defects on Printed Circuit Boards Using YOLOX Network
Generate code for a You Only Look Once X (YOLOX) object detector that can detect, localize, and classify defects in printed circuit boards (PCBs).
Deploy ONNX Student-Teacher Anomaly Detector on NVIDIA Jetson Orin Nano for Screen Defect Detection
Deploy an ONNX Student-Teacher anomaly detector on an NVIDIA® Jetson™ Orin Nano board. The workflow consists of these steps:
Deploy and Run Student-Teacher Anomaly Detection on NVIDIA Jetson
Deploy a Student-Teacher anomaly detection network as a CUDA® executable on NVIDIA® Jetson™ hardware. In this example, you generate a CUDA executable for the anomaly detection network, deploys it to the NVIDIA Jetson board, and displays the detection output on the target display.
Verify Student-Teacher Anomaly Detection Deployment Using Processor-in-the-Loop on NVIDIA Jetson
Deploy a pretrained Student-Teacher anomaly detector to NVIDIA® Jetson™ hardware and verify deployment correctness using a processor-in-the-loop (PIL) workflow. PIL execution enables you to compare the anomaly detection results produced on the target hardware with those from the host machine, verifying that the model behaves consistently after deployment.
Detect Objects Using YOLOX-Tiny Network Deployed to FPGA
Deploy a trained You Only Look Once X-tiny object detector (YOLOX-Tiny) to a target FPGA board.
Generate Code for Simulink Model that Detect Defects on Printed Circuit Boards Using YOLOX Network
Generate CUDA code for a Simulink model that detects, localizes, and classifies defects on printed circuit boards (PCBs).
