Harpia

Harpia is a visual environment (block diagrams) for evaluation, creation and implementation of image processing algorithms or computer vision systems using OpenCV.

Specific Objectives

1 General planning of the system
2 Software and graphic interface planning
3 Remote comunication planning (web management)
4 Image processing algorithms and vision tools research
5 Development of integration modules for hardware devices
6 Creation of and interface to use image processing algorithms
7 Graphic interface development
8 Development of an official webpage for the project
9 Samples application development
10 Creation of a simple tutorial about image processing
11 Integration and test os modules
12 Announcement of obtained results
13 Presentation of final results of the project
14 Development of a tutorial for learning and training of vision systems.

Reduced justification of the project

The quality control and also the obtention of confiable data inside production process is a increasing necessity inside companies of several different sectors. The occurance of errors into production processes can generate great losts and even financial problems arising from judicial actions. The industrial vision systems has as one objective the analisys of production data in a way that it is prossible the grant that defective products do not procede inside the following processes. However, even that vision systems has the capability to prevent and eliminate defective products, sometimes they are not the best solution because of the higher price, difficult implementation, hard configuration and installation.

Typical step inside development process of vision systems can be listed as: implementation of computer applications, communication and integration by software components, tests of the installed system and also its configuration and finally put the system to run definitely. This steps are based on the experience of the technical responsible which are required to have a huge specific knowledge of this are of expertise. Even with a good team of engineers and technical people, problems just appear and difficulties must be overcome.

To supply this kind of difficulties it is proposed the development of an standart environment for learning, implementation and management of vision systems under Open-Source policy.

The purpose of this project is build a graphic tool that is able to handle blocks diagrams where each task inside the final system would be a block, which are totally configurable acording to the individual necessities. Others modules that can access hardware configuration, communication with devices cameras will also be provided in further versions.

Detailed justification

The idea of development came up based on a market necessity identified by s2i group which was basically related to the cost reduction in the development of industrial vision systems. The high cost to implement a professional vision system is sometimes directly coupled with the high cost of use hardwares (cameras, lenses, ilumination devices, sensors, dedicated computers among others) and also for the required necessity to maintain a multidisciplinary group specialized in this kind of technology for research, software development

The Harpia intends to build a graphic tool with a canvas where it is possible handle blocks and configurate them to access camera devices, take photo, analise it, process it acording to a logic and finally emit a signal or not to the output of the PC. With this kind of tool, it is possible to create more quickly industrial and academic prototypes of vision systems that are able to execute some analisys over input images or video.

Expected results

1 Development of a simple, useful graphic interface for vision systems.
2 Development of a system for communication and monitoring of remote vision systems also based on high level graphic interface.
3 Development of an environment for artificial intelligence learning based on computacional vision.
4 Build an Open-Source and national system allowing the spread of knowledge inside the area of computacional vision.
5 Allow the access to a high level quality control system based on computacional vision.

Requirements:

• Python