CPU, GPU and FPGA Implementations of MALD: Ceramic Tile Surface Defects Detection Algorithm

摘要

U ovom radu razmatra se prilagodba, implementacija I usporedba performansi metode pomičnog usredn-javanja s lokalnom diferencijom (MALD) s primjenom u otkrivanju površinskih nedostataka na keramičkim pločicama. Proizvodna linija keramičkih pločica je autonomna sve do zadnje faze u kojoj je potreban ljudski vid kako bi se otkrili eventualni nedostaci na keramičkim pločicama. Nedavnim razvojem računalnih platformi I razvojem metoda računalnog vida omogućena je implementacija MALD metode na nekoliko načina. U nastojanju skraćenja vremena potrebnog za proizvodnju keramičkih pločica, MALD metoda je implementirana u trima različitim platformama: CPU (central processing unit), GPU (graphic processing unit) I FPGA (field programmable gate array), te s barem dva različita algoritma. Implementacija je izvršena sa MATLAB MEX/C++, C++, CUDA/C++, VHDL te Asembler programskim jezicima. Izmjerena vremena obrade su međusobno uspoređena za različite algoritme I njihove implementacije na različitim računalnim platformama.%This paper addresses adjustments, implementation and performance comparison of the Moving Average with Local Difference (MALD) method for ceramic tile surface defects detection. Ceramic tile production process is completely autonomous, except the final stage where human eye is required for defects detection. Recent computational platform development and advances in machine vision provides us with several options for MALD algorithm implementation. In order to exploit the shortest execution time for ceramic tile production process, the MALD method is implemented on three different platforms: CPU, GPU and FPGA, and it is implemented on each platform in at least two ways. Implementations are done in MATLAB's MEX/C++, C++, CUDA/C++, VHDL and Assembly programming languages. Execution times are measured and compared for different algorithms and their implementations on different computational platforms.