The motion characteristics of the circuit breaker are the indicator reflecting its performance. In view of the deficiencies of the existing circuit breaker motion characteristics detection technology

such as cumbersome operational steps and low measurement accuracy

this paper proposes a new circuit breaker motion characteristics detection technology based on binocular vision. Two high-speed industrial cameras are used to capture motion images of the circuit breaker connecting rod during the closing and opening process. By using the proposed parallel path search method

the normalized cross-correlation (NCC) algorithm performs frame-by-frame matching and extraction of key regions

generating images containing only critical areas. Stereo matching is performed exclusively on these key areas to obtain disparity maps which can be converted into the three-dimensional motion trajectory of the breaker connecting rod. Then 3D trajectory is converted to motion characteristic curves and parameters by coordinate transform. The test results show that

the total time per frame is reduced by 55% via parallel path search

the average error of binocular vision is 2.2% when measuring the displacement of circuit breaker connecting rod

the average deviation in depths of connecting rod during the breaker closing and opening is 2.24 mm and −2.65 mm

respectively

the opening displacements in X and Y directions are 24.528 mm and 1.365 mm

and the average speeds are 0.947 m/s and 0.045 m/s

respectively. In summary

circuit breaker motion characteristic detection technology based on binocular vision achieves accurate non-contact measurement of three-dimensional motion characteristics

and this approach simplifies procedure of measurement and improve the efficiency of analysis and calculation.