Vpliv barve in jakosti svetlobe na prepoznavo obdelovanca s strojnim vidom

VIRANT, MARCEL (avtor), Mihelj, Matjaž (mentor), Perš, Janez (komentor)

Zaključna naloga obravnava raziskavo vpliva barve in jakosti svetlobe na prepoznavo obdelovanca s strojnim vidom. Opisane so teoretične osnove strojnega vida, digitalne slike, osvetlitve in zaznave predmeta, barve svetlobe ter zajema slike. V osrednjem delu diplomskega dela je opisan postopek izvedbe študije. Študija temelji na medsebojni odvisnosti štirih različnih veličin, oddaljenosti kamere od obdelovanca, barvi osvetlitve, jakosti osvetlitve ter materialu obdelovanca. Rezultati meritev so predstavljeni v preglednicah, iz katerih je nato preračunana povprečna vrednost prepoznave pri določeni barvi in jakosti svetlobe. Rezultati študije so pokazali, da je bila v večini primerov najustreznejša bela barva. To sicer ne pomeni, da ostale barve niso dosegale dobrih rezultatov. V veliko primerih bi lahko namesto bele barve uporabili tudi zeleno ali modro, saj sta bili prav tako ustrezni kot bela. Rdeča barva je bila v nekaterih primerih bolj učinkovita kot modra ali zelena, v večini primerov pa je bila neuporabna.

• Jezik: Slovenski jezik
• Ključne besede: strojni vid, jakost in barva svetlobe, RGB svetloba, prepoznava obdelovanca.
• Vrsta gradiva: Diplomsko delo/naloga
• Organizacija: FE - Fakulteta za elektrotehniko
• Leto izida: 2020
• PID: 20.500.12556/RUL-119998
• Datum objave v RUL: 15.09.2020

Sekundarni jezik

• Jezik: Angleški jezik
• Naslov: Influence of colour and light intensity on workpiece recognition with machine vision

Izvleček

This thesis deals with the research of the influence of colour and light intensity on the recognition of a workpiece with machine vision. The theoretical foundations of machine vision, digital images, illumination and perception of an object, light colours and the image acquisition are described. The central part of the diploma thesis describes the study procedures. The study is based on the interdependence of four variables, the distance of the camera from the workpiece, the colour of the illumination, the intensity of the illumination and the material of the workpiece. The results of the measurements are presented in tables, from which the average value of recognition at a certain colour and light intensity is calculated. The results of the study showed that in most cases, white was an adequate colour. This does not mean that other colours did not achieve good results. In many cases, also green or blue were just as useful as white. Red was in only a few cases more appropriate than blue or green, and in most cases, it was not adequate.

• Ključne besede: machine vision, colour and the intensity of light, RGB light, recognition of the workpiece.