Embedded Microcontrollers Benchmarking using Sliding Window Algorithm
Benchmarking using standard signal processing algorithms is very convenient to compare different embedded microcontrollers performance but is not always enough to answer the question if the particular microcontrollers will meet real-time requirements of the considered task. Sliding window algorithm used in materials surface quality visual inspection is described. Speed of this algorithm execution was tested on three common 8, 16 and 32 bit microcontrollers (Atmega16, MSP430F449 and ARM7 core based STR712F). Influence of different implementation versions of the algorithm in C language as well as compiler optimization settings upon execution speed were investigated. Applied approach enables to select the best algorithm implementation in order to optimize execution speed. Atmega16 microcontroller exhibited the fastest operation set 4 MHz clock frequency for all microcontrollers. However, STR712F featured the highest throughput if every microcontroller is set to operate at its top CPU clock frequency. Ill. 4, bibl. 5 (in English; summaries in English, Russian and Lithuanian).
Copyright terms are indicated in the Republic of Lithuania Law on Copyright and Related Rights, Articles 4-37.