Sensor arrays are widely used to extract the distribution of variables in large sensing areas or collect multi-degree-of-freedom data. In this paper, we propose adaptive resistance matrix approaches, which can get rid of additional electrical components such as diodes, transistors, multiplexers, op-amps, switches, and current sources. By eliminating the power supply parameters, these approaches can show great advantages and application potential when they are used to measure wearable flexible sensor arrays or other low-cost sensor arrays. Experimental results have shown that the proposed approaches achieved a high accuracy (99.45% and 99.80%), large measurement range (0.5 ∼ 1.5 times the reference resistance value), and relatively low crosstalk error (0.066% and 0.128%). The approach was then applied to a flexible strain sensor array to estimate the three-degree-of-freedom lumbar spine motion, and the results have shown that the resistance distribution of the sensor array can clearly show the strain distribution under different lumbar movements.