MODUL SOFTWARE PENTRU LOCALIZAREA SI RECUNOASTEREA FORMEI OBIECTELOR IN VEDEREA PREHENSIUNII ANTROPOMORFE

Autor/autori: Dr. ing. Alexandru Mihail ITU, Prof. univ. dr. ing. Eur Ing Ionel STARETU

Rezumat: In aceasa lucrare este prezentata structura generala, particularitatile implementarii si evaluarea eficientei unui modul software pentru localizarea, vizualizarea, recunoasterea formei, masurarea si generarea modelului 3D ale unor obiecte destinate prehensiunii cu un prehensor antropomorf . Vizualizarea se face cu o camera video performanta, iar recunosterea formei se face cu metoda clsificatorilor. Dupa stabilirea formei obiectului, prehensorul se va apropia de tinta, deplasarea acestuia fiind masurata. Pe baza masurarii distantei dintre punctul initial si un al doilea punct intermediar, se pot face calcule ce vizeaza dimensiunea obiectului si distanta ramasa pana la acesta. Implementarea si modul de lucru cu software-ul conceput se face in mai multi pasi care sunt precizati si descrisi. Evaluarea performantelor sistemului software se face in functie de o serie de factori externi, cei mai importanti fiind: conditiile de lumina; calitatea dispozitivului video; dimensiunea capturii; factorul uman, printr-o serie de experimente care au vizat capacitatea sistemului de a face fata factorilor externi mai sus enuntati.

Cuvinte cheie: procesare imagine, recunoasterea formei, modul software, prehensiune antropomorfa

Abstract: This paper presents the general structure, implementation features and a software module for tracing, visualization, shape recognition, measurement and efficiency evaluation for 3D model generation of objects to grip by an anthropomorphic gripper. Viewing is possible with an advanced video camera and shape recognition is possible through classifiers method. After setting the object shape, the gripper will approach the target, and its displacement is measured. Based on measurements of the distance between the initial point and a second intermediate point, you can do calculations aimed at object size and remaining distance to it. Implementation, and how to work with the software designed are possible in several steps that are specified and described. The performance evaluation of the software system is based on a series of external factors, the most important being: light conditions, quality of video device, capture size, the human factor, through a series of experiments focused on the system's ability to cope with external factors above mentioned.

Keywords: image processing, shape recognition, software module, anthropomorphic gripping

DOWNLOAD PDF