CKVision SDK 6.0 | Machine Vision SDK Overview & Use Cases

1. Overview

CKVision SDK 6.0 is a professional machine vision development toolkit for the industrial vision inspection field. The SDK provides complete functional modules including 2D/3D image processing, object detection, precision measurement, defect detection, and intelligent recognition. It can be widely used in industrial automation scenarios such as intelligent manufacturing, quality inspection, and robotic vision guidance.

Core Features

• Modular architecture: Adopts DLL component-based design, with independently encapsulated functional modules for flexible integration
• High-performance algorithms: Optimized image processing and analysis algorithms to meet real-time inspection requirements
• 2D/3D fusion: Supports both planar image and 3D point cloud data processing
• Cross-platform support: Provides standard C++ API and Qt graphical interface
• Industrial-grade stability: Optimized for industrial environments with high reliability and robustness

2. Functional Module Architecture

2.1 Core Basic Module (CKBase.dll)

The basic module is the core of the entire SDK, providing underlying data structures and general image processing functions.

Image Storage Structures

• CKImage: Standard 2D image object supporting multiple pixel formats
• CKMask: Image mask for ROI (Region of Interest) definition
• CK3DImage: 3D point cloud image data structure
• CKSurface: 3D surface data representation
• CKMaskImage: Mask composite image for complex region processing

Image Processing Tools

• CKImgOpera: Arithmetic and logical operations (add, subtract, multiply, divide, AND, OR, NOT, etc.)
• CKImgFilter: Filtering (mean, Gaussian, median, bilateral filtering, etc.)
• CKImgMorph: Morphological operations (erosion, dilation, opening, closing, etc.)
• CKImgTrans: Geometric transformation (rotation, scaling, affine, perspective transformation)
• CKImgConve: Image conversion (format conversion, color space conversion)

Image Analysis Tools

• CKHistogram: Histogram statistics and grayscale analysis
• CKPixelStat: Pixel-level statistical analysis
• CKProjection: Image projection curve generation
• CKSharpAssess: Sharpness evaluation (for auto-focusing)

Calibration Functions

• CKDotMatrix: Dot pattern calibration board detection
• CKChessboard: Chessboard calibration board detection
• CKImageCalib: Image distortion correction
• CKPointCalib: Coordinate system calibration and transformation

Auxiliary Tools

• CKMat2D: 2D matrix operations
• CKGeometry: Geometric measurement (point-line distance, angle, etc.)
• CKFrameTrans: Coordinate system transformation
• CKImgPyramid: Image pyramid (multi-scale processing)

Typical Application Scenarios

• Image preprocessing: denoising, enhancement, binarization
• Camera calibration: eliminate lens distortion, establish pixel-to-physical dimension mapping
• Image registration: alignment of multi-camera or multi-view images

2.2 Precision Measurement Module (CKGauge.dll)

Designed for high-precision geometric measurement, providing sub-pixel edge detection and shape fitting.

Core Tools

• CKEdgeTool: Edge detection with sub-pixel localization
• CKSpaceTool: Distance measurement
• CKScanEdge: Edge point scanning along a path
• CKScanSpace: Distance scanning along a path
• CKFitCircle: Circle fitting (least squares method)
• CKFitLine: Line fitting
• CKFitEllipse: Ellipse fitting

Measurement Accuracy

• Edge localization accuracy up to 0.1 pixel
• Supports caliper tools, multi-point fitting, and other measurement modes

Application Scenarios

• PCB pad dimension inspection
• Bearing inner/outer diameter measurement
• Part contour dimension verification
• Assembly gap measurement
• Geometric parameter inspection of precision mechanical parts

2.3 Defect Detection Module (CKDetect.dll)

Provides various defect detection algorithms for surface quality inspection and anomaly recognition.

Functional Components

• CKPatInspect: Image comparison inspection (template-matching defect detection)
• CKBlobAnalyzer: Blob analysis (connected region analysis)

• Feature extraction: area, perimeter, roundness, rectangularity, etc.
• Defect classification and filtering
  • CKEdgeContour: Contour detection (edge extraction and contour tracking)
  • CKRidgeContour: Contour centerline extraction
  • CKContourDefect: Contour defect detection (dents, bulges, breaks, etc.)
  • CKCharFinder: OCV character segmentation (Optical Character Verification)

Detection Capabilities

• Surface defects: scratches, stains, bubbles, pits
• Assembly defects: missing, redundant, misaligned parts
• Out-of-tolerance dimensions, shape anomalies

Application Scenarios

• Panel defect inspection (display, glass, etc.)
• Print quality inspection
• Metal surface scratch detection
• Tablet integrity inspection
• Welding quality inspection

2.4 Location and Recognition Module (CKLocate.dll)

Enables fast and accurate object location and pattern recognition.

Core Algorithms

• CKContourMatch: Contour matching

• Fast shape-based matching
• Robust to rotation, scaling, and occlusion
  • CKModelContour: Model contour management
• Template training and storage
• Multi-template management

Performance Indicators

• Matching accuracy: ±0.1 pixel (position), ±0.1° (angle)
• Search speed: millisecond-level (typical scenarios)

Application Scenarios

• Robotic vision guidance (grasping and positioning)
• Product sorting (pose recognition)
• Assembly positioning (part alignment)
• Label positioning
• Irregular object recognition

2.5 Code Recognition Module (CKReader.dll)

Supports reading and localization of various 1D and 2D codes.

Supported Codes

• CKBarcode: 1D barcode reading

• Code 39, Code 128, EAN-13, UPC, etc.
  • CKReadQRCode: QR code reading
  • CKDataMatrix: DataMatrix code reading
  • CKFindBarcode: Barcode localization (no decoding, only positioning)
  • CKReadOcr: Character recognition (OCR)

Technical Features

• High fault tolerance: adapts to dirty, blurred, skewed conditions
• Fast decoding: optimized decoding algorithms
• Batch recognition: multiple codes in one image

Application Scenarios

• Logistics sorting (express number recognition)
• Product traceability (QR code reading)
• Quality management (batch number recognition)
• Warehouse management (shelf code recognition)
• Electronic product serial number reading

2.6 Color Processing Module (CKColor.dll)

Provides color analysis and color segmentation functions.

Functional Components

• CKColorMonitor: Color monitoring (color consistency inspection)
• CKColorIdentify: Color recognition (color classification)
• CKRGBThreshold: Color extraction in RGB space
• CKHSVThreshold: Color extraction in HSV space (closer to human perception)

Application Scenarios

• Color sorting (fruits, tablets, etc.)
• Printing color difference inspection
• Liquid color detection
• Cable identification (wire harness color distinction)
• LED color inspection

2.7 3D Measurement Module (CK3DTool.dll)

Processes and analyzes 3D point cloud data.

Functional Components

• CK3dGeom: 3D geometric measurement (distance, angle, volume)
• CK3dHeight: Height detection (Z-axis measurement)
• CK3dPlane: Plane fitting (flatness inspection)
• CK3dSphere: Sphere fitting (sphericity inspection)
• CK3dRidge: Peak detection
• CK3dAnalyse: Height analysis (height distribution statistics)
• CK3dVolume: Volume measurement
• CK3dFilter: 3D filtering (point cloud denoising)
• CK3dEdge: 3D edge measurement

3D Data Sources

• Structured light camera
• Laser scanner
• TOF camera
• Binocular stereo vision

Application Scenarios

• Weld seam height inspection
• Glue coating thickness measurement
• Workpiece volume measurement
• Flatness/sphericity inspection
• 3D defect detection (pit depth, etc.)
• Material accumulation volume measurement

2.8 Graphics Display Module (CKGDI.dll / CKQtGDI.dll)

Provides visual graphics drawing for result display and interactive debugging.

Drawing Elements

• CKGdiView: View management
• CKGdiText: Text annotation
• CKGdiPoint: Point drawing
• CKGdiLine/CKGdiLineScan: Line and scan line
• CKGdiRect/CKGdiRotBox: Rectangle (orthogonal/rotated)
• CKGdiCircle/CKGdiArc/CKGdiRing: Circle, arc, ring
• CKGdiEllipse: Ellipse
• CKGdiContour/CKGdiPolygon: Contour, polygon
• CKGdiBoxScan/CKGdiRingScan/CKGdiPolyScan: Various scan regions
• CKGdiProfile: Profile line display
• CKGdiHistogram: Histogram display
• CKGdiMask: Mask display
• CKGdiFrame: Coordinate system display (CKGDI only)

Dual Implementation

• dll: Windows GDI-based implementation
• dll: Qt-based cross-platform implementation

Application Value

• Visual debugging: intuitively view inspection results
• Human-machine interaction: ROI setting, parameter adjustment
• Report generation: annotation and output of inspection results

3. Typical Application Cases

3.1 Electronic Manufacturing

PCB Solder Joint InspectionWorkflow:Camera capture → Image enhancement (CKImgFilter) → Contour detection (CKEdgeContour)→ Circle fitting (CKFitCircle) → Defect judgment (CKBlobAnalyzer)

Inspection Items

• Solder joint diameter (CKGauge measurement)
• Solder joint roundness (geometric feature analysis)
• Insufficient/excess solder (area statistics)
• Cold solder/false solder (grayscale analysis)

Component Mounting InspectionWorkflow:Image capture → Template matching (CKContourMatch) → Pose recognition→ Position offset measurement (CKEdgeTool) → Polarity inspection (CKPatInspect)

Inspection Items

• Component position offset
• Angle deflection
• Front/back judgment
• Missing/wrong components

3.2 Automotive Manufacturing

Body Weld Seam Inspection (3D)Workflow:3D scanning → Point cloud filtering (CK3dFilter) → Edge extraction (CK3dEdge)→ Height measurement (CK3dHeight) → Weld quality judgment

Inspection Items

• Weld height (0.1mm accuracy)
• Weld width
• Weld continuity
• Porosity defects

Part Dimension InspectionToolchain:CKImageCalib (camera calibration) → CKEdgeTool (edge localization)→ CKFitLine/CKFitCircle (geometric fitting) → CKGeometry (dimension calculation)

Applications: Bearing inner/outer diameter, shaft diameter, bolt length, etc.

3.3 Pharmaceutical Industry

Tablet Quality InspectionWorkflow:Backlight imaging → Binarization (CKHistogram) → Blob analysis (CKBlobAnalyzer)→ Defect detection → Color inspection (CKColorIdentify)

Inspection Items

• Breakage (contour defect)
• Stains (Blob area)
• Color anomaly
• Shape deformation

Medicine Bottle Label InspectionOCR recognition:Image preprocessing → Character segmentation (CKCharFinder) → OCR recognition (CKReadOcr)

Barcode reading:CKBarcode/CKReadQRCode → Batch number verification

3.4 Food and Beverage Industry

Package Inspection

• Label positioning: CKContourMatch (logo recognition)
• Print inspection: CKPatInspect (comparison with standard template)
• Date code reading: CKReadOcr (production date recognition)
• Sealing inspection: CKEdgeContour + CKContourDefect (sealing integrity)

Liquid Level InspectionWorkflow:Side imaging → Edge detection (CKScanEdge) → Liquid level localization→ Height calculation → Pass/fail judgment

Assisted by projection curve (CKProjection) for liquid level recognition.

3.5 Logistics and Warehousing

Express SortingSystem architecture:High-speed camera capture → Waybill localization (CKFindBarcode)→ 1D/2D code decoding (CKBarcode/CKReadQRCode)→ Sorting control signal output

Performance Requirements

• Recognition speed: <100ms
• Recognition rate: >99.5%
• Supports multiple codes and damaged codes

3.6 Textile and Garment Industry

Fabric Defect InspectionWorkflow:Line-scan camera capture → Image stitching → Defect detection (CKPatInspect)→ Blob analysis (CKBlobAnalyzer) → Defect classification

Inspection Content

• Holes, stains, loose threads
• Color difference, stripes
• Texture anomalies

4. Development and Integration Guide

4.1 System Requirements

Operating System

• Windows 7/8/10/11 (x64)
• Linux (Ubuntu 18.04+)

Development Environment

• Visual Studio 2015 or later (Windows)
• GCC 7.0+ (Linux)
• Qt 5.x (if using CKQtGDI)

Hardware Requirements

• CPU: Intel i5 or above
• RAM: 8GB+ (16GB recommended)
• Discrete GPU recommended (for image display)

4.2 Quick Start

Basic Initialization Flow

cpp

运行

#include “CKBase.h”#include “CKDetect.h”#include “CKGauge.h”

// 1. Initialize CKVision libraryCKVision::InitLibrary();

// 2. Create image object

CKVision::CKImage image;

image.Create(1920, 1080, 8);  // 1920×1080, 8-bit grayscale

// 3. Load image

image.LoadFromFile(“test.bmp”);

// 4. Use functional modules// … Perform inspection/measurement …

// 5. Exit libraryCKVision::ExitLibrary();

Typical Measurement Example

cpp

运行

#include “CKGauge/CKEdgeTool.h”#include “CKGauge/CKFitCircle.h”

// Create edge detection tool

CKVision::CKEdgeTool edgeTool;

edgeTool.SetImage(image);

edgeTool.SetROI(centerX, centerY, width, height, angle);

edgeTool.SetEdgeThreshold(30);

edgeTool.SetEdgePolarity(CK_EDGE_POSITIVE);

// Execute edge detectionif (edgeTool.Execute()){

// Get edge points

std::vector<CKPoint2f> edgePoints = edgeTool.GetEdgePoints();

 

// Circle fitting

CKVision::CKFitCircle fitCircle;

fitCircle.SetPoints(edgePoints);

fitCircle.Execute();

 

// Get results

float centerX = fitCircle.GetCenterX();

float centerY = fitCircle.GetCenterY();

float radius = fitCircle.GetRadius();

float error = fitCircle.GetFitError();}

4.3 Performance Optimization Suggestions

Image Preprocessing Optimization

• Use ROI to limit processing area and reduce computation
• Select appropriate filter size (oversize reduces speed)
• Use image pyramid for multi-scale processing

Algorithm Selection

• Prefer fast algorithms for simple scenes (Blob analysis > deep learning)
• Combine multiple algorithms for complex scenes (coarse location → fine measurement)

Parallel Processing

• Multi-threading for multi-camera systems
• GPU acceleration for filtering and computation

Memory Management

• Release unused image objects promptly
• Avoid frequent creation/destruction of large images

5. Technical Support and Services

5.1 Document Resources

• API reference manual (detailed function descriptions)
• Algorithm principle whitepaper
• Application case documents
• Frequently Asked Questions (FAQ)

5.2 Development Support

• Sample code library (covers all modules)
• Technical forum
• Online technical support
• Custom development services

5.3 Training Services

• Basic training: SDK getting started
• Advanced training: algorithm principles and tuning
• Industry application training: industry-specific solutions

6. Summary and Outlook

CKVision SDK 6.0, with its complete functional modules, high-performance algorithms, and industrial-grade stability, has become an ideal machine vision development toolkit. It provides reliable solutions for both 2D planar inspection and 3D volumetric measurement, from simple dimension measurement to complex defect recognition.

Core Advantages

• Completeness: covers full vision inspection flow (capture → process → analyze → output)
• Modularity: on-demand integration, reduces development complexity
• High precision: sub-pixel measurement meets high-precision demands
• Usability: clear API design, rich sample code
• Industry adaptability: optimized for electronics, automotive, pharmaceutical, etc.

Future Development Directions

• Deep learning integration (CNN defect detection, AI classification, etc.)
• Enhanced 3D capabilities (point cloud registration, 3D reconstruction)
• Edge computing optimization (ARM/embedded support)
• Cloud collaboration (cloud-based vision data analysis)

Choosing CKVision SDK 6.0 means choosing a mature, reliable, and continuously evolving machine vision development platform, helping your intelligent manufacturing projects launch quickly and run stably long-term.