This page uses cookies. For details and possible settings refer to our Privacy Policy.
Continuing to use this page means accepting the processing of cookie files.

Accept (hide info)
POL ENG GER 中文(繁體)

Deep Learning Add-on

Introduction

Deep Learning Add-on is a new breakthrough in machine vision applications. It is a set of ready-made tools which are trained with Good and Bad samples, and which then detect defects or features automatically. Internally it uses large neural network structures, designed and optimized by our research team for use in industrial inspection systems. For the user, however, they are provided as simple filters with very few parameters, and with easy-to-use graphical tools for convenient execution of the training process.

Deep Learning Add-on truly embodies the main principles of Adaptive Vision:

Intuitive - can be used even by users with no programming skills

Powerful - unleashing advanced capabilities of neural networks

Adaptable - deep learning models can be retrained to include new features.

Key Facts

Training Data

Typical applications require between 20 and 50 images for training.

Hardware

A modern GPU is recommeded for fast training and execution.

Speed

Assuming use of a GPU, typical training time is 5 minutes and typical execution time is 100ms/MP.

Deep Learning vs Traditional Machine Vision

Deep Learning is a new reliable solution for machine vision problems that could not have been solved before. There are, however, applications that still can only be realized with traditional methods. How do you know, which approach is better? Here is a quick guide:

Deep Learning

Traditional Machine Vision

Typical applications:

    • Surface inspection (cracks, scratches)
    • Food, plant, wood inspection
    • Plastics, injection moulding
    • Textile inspection
    • Medical imaging

Typical applications:

    • Dimensional measurements
    • Code reading
    • Presence or absence checking
    • Robot guidance
    • Print inspection

Typical characteristics:

    • Deformable objects
    • Variable orientation
    • Customer provides vague specification with examples of Good and Bad parts
    • Reliability 99%

Typical characteristics:

    • Rigid objects
    • Fixed orientation
    • Customer provides formal specification with tolerances
    • Reliability 100%

Training Procedure

1. Collect and normalize images

    • Acquire between 20 and 50 images, both Good and Bad, representing all possible object variations; save them to disk
    • Make sure that the object scale, orientation and lighting are as consistent as possible
  • 2. Training

    • Use a tool from Deep Learning Add-on
    • Open Deep Learning editor
    • Load training images
    • Label images as Good or Bad (semi-unsupervised mode), or mark masks with drawing tools (supervised mode)
    • Click “Train”
  • Training and Validation Sets

    In Deep Learning, as in all fields of machine learning, it is very important to follow correct methodology. The most important rule is to separate the Training set from the Validation set. The Training set is a set of samples used for creating a model. We cannot use it to measure the model’s performance, as this often generates results that are overoptimistic. Thus, we use separate data – the Validation set – to evaluate the model. Our Deep Learning tool automatically creates both sets from the samples provided by the user.

    3. Execute

    • Run the program and see the results
  • Feature Detection

    In the supervised mode the user needs to carefully label pixels corresponding to defects on the training images. The tool then learns to distinguish good and bad features by looking for their key characteristics.

    Application Examples

    Photovoltaics Inspection

    In this application cracks and scratches must be detected on a surface that includes complicated features. With traditional methods, this requires complicated algorithms with dozens of parameters which must be adjusted for each type of solar panel. With Deep Learning, it is enough to train the system in the supervised mode, using just one tool.

    Satellite Image Segmentation

    Satellite images are difficult to analyse as they include a huge variety of features. Nevertheless, our Deep Learning Add-on can be trained to detect roads and buildings with very high reliability. Training may be performed using only one properly labeled image, and the results can be verified immediately. Add more samples to increase the robustness of the model.

    Anomaly Detection

    In the unsupervised mode training is simpler. There is no direct definition of a defect – the tool is trained with Good samples and then looks for deviations of any kind.

    Application Examples

    Package Verification

    When a sushi box is delivered to a market, each of the elements must be correctly placed at a specific position. Defects are difficult to define when correct objects may also vary. The solution is to use unsupervised deep learning mode that detects any significant variations from what the tool has seen and learned in the training phase.

    Plastics, injection moulding

    Injection moulding is a complex process with many possible production problems. Plastic objects may also include some bending or other shape deviations that are acceptable for the customer. Our Deep Learning Add-on can learn all acceptable deviations from the provided samples and then detect anomalies of any type when running on the production line.

    Object Classification

    The Object Classification tool divides input images into groups created by the user according to their particular features. As a result the name of a class and the classification confidence are given.

    Application Examples

    Caps: Front or Back

    Plastic caps may sometimes accidently flip in the production machine. The customer wants to detect this situation. The task can be completed with traditional methods, but requires an expert to design a specific algorithm for this application. On the other hand we can use deep learning based classification which automatically learns to recognize Front and Back from a set of training pictures.

    3D Alloy Wheel Identification

    There may be hundreds of different alloy wheel types being manufactured at a single plant. Identification of a particular model with such quantities of models is virtually impossible with traditional methods. Template Matching would need huge amount of time trying to match hundreds of models while handcrafting of bespoke models would simply require too much development and maintenance. Deep learning comes as an ideal solution that learns directly from sample pictures without any bespoke development.

    Instance Segmentation

    The instance segmentation technique is used to locate, segment and classify single or multiple objects within an image. Unlike the feature detection technique, this technique detects individual objects and may be able to separate them even if they touch or overlap.

    Application Examples

    Nuts Segmentation

    Mixed nuts are a very popular snack food consisting of various types of nuts. As the percentage composition of nuts in a package shall be in accordance with the list of ingredients printed on the package, the customers want to be sure that the proper amount of nuts of each type is going to be packaged. Instance segmentation tool is an ideal solution in such application, since it returns masks corresponding to the segmented objects.

    Package Verification

    A typical set of soup greens used in Europe is packaged on a white plastic plate in a random position. Production line workers may sometimes accidently forget to put one of the vegetables on the plate. Although there is a system that weighs the plates, the customer wants to verify completeness of the product just before the sealing process. As there are no two vegetables that look the same, the solution is to use deep learning-based segmentation. In the training phase, the customer just has to mark regions corresponding to vegetables.

    Point Location

    The Point Location tool looks for specific shapes, features or marks that can be identified as points in an input image. It may be compared to traditional template matching, but here the tool is trained with multiple samples and becomes robust against huge variability of the objects of interest.

    Application Examples

    Bees Tracing

    The task that seems impossible to achieve with traditional methods of image processing can be done with our latest tool. In this case we use it to detect bees. When it is done we can check whether they are infected by varroosis – the disease caused by the parasitic mites attacking the honey bees. The parasite attaches to their bodies and upon the basis of a characteristic red inflammation we can classify them according to their health condition. Not only does this example show that it is an easy solution for a complex task, but also that we are open to many different branches of industry e.g. agriculture.

    Pasta

    Tagliolini is a type of pasta cut into long strips about 3 mm wide and formed into portions of a characteristic shape. After production process the portions are packed into boxes and before delivering them to market, the customer wants to verify whether the number of portions is correct. This task is challenging as it is difficult, sometimes even for a human, to define the boundaries of the portions. The solution is a deep learning-based point location tool which automatically learns to locate all the instances of objects pointed in the training phase.