AT logo - The Machine Vision Specialists  
  Frame Grabbers  
  Vision Systems  
  Contact AT  
  Site Map  

Machine Vision Newsletter


(08) 9242 5411

(02) 9905 5551

(03) 9384 1775

Defence Recognised Supplier Scheme Logo

MV Tools
Back to Software
Back to Manufactuers


MVTools software, the core vision algorithm library within Sherlock, is a complete C/C++ library of highly robust grayscale machine vision tools. Unique to MVTools is Edge Crawler, a new geometric pattern-finding tool which extracts and reports contour points in subpixel accuracy. Some of the additional algorithms available are grayscale Search, optional SMART SearchT, blob, morphology, point processing, convolution, Fourier transform, non-linear area processing, geometric transformation, color image processing, camera and display abstractions, and so much more for automation professionals developing and deploying high-performance machine vision within high-volume OEM products.

Designed for extreme speed, accuracy, and ease-of-use, MVTools offers the most comprehensive set of robust vision tools that speeds time-to-market for OEMs, system integrators, and automated machine builders. A hardware independent package, MVTools now supports any vendors' frame grabbers that deliver 8- or 16-bit data to memory.

Once you have proven your application's performance from within MVTools development environment - Sherlock -- and are prepared to deploy your application, several options are available to the OEM or system integrator. For OEMs, who are commonly concerned with embedding the most reliable and cost-effective solution, MVTools C/C++ library or DLLs allow OEMs to write their own VisualBasic or Visual C++ front ends to access MVTools machine vision algorithms. The economical MVTools run-time license enables OEMs to minimize their total machine vision hardware and software costs and deliver the most cost-effective solution.

Alternatively, for OEMs who would prefer to program vision tasks via OCX, users can choose to run the Sherlock application in server mode and communicate with the vision application using a VisualBasic or Visual C++ front end.


Features & Benefits

  • Robust and proven software library of core machine vision routines - Lets developers focus on their specific competitive issues while providing speed and accuracy normally found only in intensively engineered solutions.
  • Use the Windows environment of Sherlock as a prototyping tool - Quickly determine the performance of each algorithm before writing the first line of applications code
  • Designed to fully exploit Intel MMX technology and advanced optimization techniques - Many MVTools routines are available in both C++ and MMX calls.
  • Edge Crawler - A fundamental tool in extracting high level features from objects such as straight lines, curves, circles and curvature points.
  • Subpixel accuracy - Measurements returned have been fully tested to assure you they are the most accurate in the industry.
  • Supports any vendors' frame grabbers that deliver 8- or 16-bit data to memory
  • Supports floating point operations


  • Alignment
  • Gauging
  • Flaw Detection
  • Presence/Absence
  • OCV (Optical Character Verification)
  • Robot Guidance
Development Cycle

Whether you have a high or low volume application, use Sherlock as your evaluation tool. Using Sherlock, you can completely prototype the application without writing a single line of code. Once satisfied with performance, deployment can be executed in three ways.

  1. For low volume applications, you can immediatly install using the Sherlock environment without additional development time.
  2. If you want to put a custom user interface on Sherlock, you can write a Visual Basic program calling the Sherlock application from server mode.
  3. For high volume applications, you can write a complete custom application calling MVTools, the core machine vision algorithms from Sherlock. Programming using MVTools after prototyping in Sherlock assures performance before spending a lot of development time.

MVTools Structure

MvTools Components
  • Library, DLL & Header Files for Algorithms, Image Utilities and Imager for Win95/98 or WinNT 4.0/2000
  • MVTools examples in source and executable form: (5) "C", (2) "C++" , and (1) Visual Basic examples
  • On-line and printed user manual
  • Camera Configurator (hardware set-up utility)
  • Hardware driver level functions and API
  • Source code for Imager
  • Optional Dongle (sold separately) for use with other vendors' frame grabbers
Operating Systems and Hardware

MVTools runs on Microsoft Win95/98 and Win4.0/2000. If your computer has MMX technology, MVTools will automatically use it to greatly increase the speed of operations such as image subtraction for "golden template" defect detection.

MvTools Core Functions

These functions represent the basic MVTools image processing and analysis operations. These functions can be separated into two categories; Filter Functions and Measurement Functions.

Filter Functions

  • Threshold
  • Image Mathematics
  • Geometric
  • Projection and Moments
  • First and Second Differences
  • Convolution
  • Morphological
  • Image Averaging
  • Array Manipulation
  • Labeling
  • Linked list
  • Fourier Transform
  • Color Processing
  • Contour Extraction

Measurement Functions

  • Centroid
  • Measurement and Geometry
  • Locate Extrema
  • Wave Analysis
  • Lead Analysis
  • EdgeDdetection and Analysis
  • Coordinate Transform
  • Thread
  • Image Generation
  • Blob (connectivity) Analysis
  • Protrude
  • Point Operations
  • Statistical
  • Barcode 39 reading
  • Alignment
  • Pattern search
  • Optional SMART Search (sold separately)
MvTools Benchmarks

MVTools performance compares well with other machine vision libraries. An important point is that MVTools was developed from a working and highly tested machine vision package (Sherlock), while other libraries were developed as a library. This means that the routines in MVTools are what you need for real machine vision applications, not functions from a textbook or laboratory. It also means that the MVTools routine might be slower than an apparently similar routine because MVTools is more robust.


If you like this page, please recommend it and share it.

Facebook Twitter More