Intopii

Reads images (and soon videos)

Pinta learns from samples. Collect a representative set of images or videos about your problem and feed them to Trainer using Image Source options. Pretty much any format goes.

See the big picture

If you are only interested in the large scale shapes and patterns in the image, use scaling options to get rid of unnecessary details - and get performance gain as a size effect.

Segment intellegently

Segmentation using Pinta is as easy as selecting the Image Spitter and adjusting the size and overlap of the tiles as needed to achieve the desired accuracy.

Analyze structures

One of the most powerful image features in existence, the LBP texture features can make your classifier to measure complex structural differences in the image, from tiniest details to largest of scales.

Analyze brightness

Grayscale histogram can be used to make your analyzer know whether the image, or part of it, is dark, bright, or anything in between. Enough alone in many simple applications.

Analyze colours

Customizable color histograms, percentiles, and color correlograms (new!) make your analyzer accurately measure variantions in the color distribution of the image. A must when colors matter - and often they do.

Combine features (new!)

Feature combiner allows your analyzer draw conclusions based on the combination of many features at the same time, much like we humans do.

Self organizing intelligence

The Self Organizing Map (SOM) is the brain (the size, elasticity, and thoroughness of which you can determine) of your analyzer. It learns from the samples and classifies image data when trained.

Know the unknown (coming)

Reject threshold can be used to teach the analyzer recognize a single class of images. Anything outside that goes to a special class of unknowns. Useful for collecting additional training samples, for example.

Let it learn

Trainer teaches your analyzer the characteristics of your material based on the samples and selected features. You don't need to categorize samples in any way beforehand. Let Trainer do the hard work and relax.

See the result

Evaluate learning and classification results and name the automatically generated categories with an intuitive image map that practically lets you see what your analyzer thinks about the images.

Categorize as you wish

Named classes? Numerical values? Metrics allow you to freely define what kind of results you want from your analyzer.

Modify in Trainer

Analyzers can be reviewed, modified, and tuned with Trainer whenever necessary. They can also be used as templates for new analyzers and completely retrained for new purposes.

Connect to your applications

Integrate and use analyzers in your own applications through the API. They make your applications understand what they see. Feed them with image data and get results, instantly.

Distribute

As compact files that encapsulate all the skills and knowledge you have taught them, analyzers can be effortlessly distributed to and used by other users of Pinta.

Strong foundation

Pinta is built on top of Into, the most advanced pattern recognition and image analysis software platform on planet.

Learn more about Into

Tested in serious applications

Earlier versions of Pinta are being used around the world in demanding industrial applications such as real-time surface analysis of steel and paper. It is programmed to run 24/7 without a glitch.

Cross-platform compatibility

Pinta runs on Windows XP and Vista, and on different flavors of Linux, both on 64 and on 32 bits. Custom builds can be provided for embedded systems.

Built to be customized

Although Pinta works out of the box in surprisingly many cases, it is built to be enhanced with special processing stages without changing a line of code.

See application examples

Help at hand

You will not be alone with Pinta. We are ready and eager to help you in building even more complicated applications and even faster.

Learn more about support