The traditional method for determining the oil content of oily sludge was improved, and the extraction smartphone image colorimetry was established by combining the image colorimetry with the pretreatment method of hot solvent extraction. The effects of RGB value, different color channels, shutter time, sensitivity, white balance and other parameters on the determination of oil content were investigated. Under the premise that the light source was tungsten iodide lamp and the front camera of Samsung A8s mobile phone was used, the best test conditions were finally determined: white balance value was 2 800 K, sensitivity value was 100, shutter time was 1 / 2 000 s. The linear correlation coefficients of the method under the optimal test conditions were more than 0.99 under the optimal test conditions, and the linear range was similar to that of the spectrophotometric method, with the detection limit of 0.06% and the lower limit of 0.24%. The recoveries of 8 groups of parallel standard samples were 96%~103%, and the relative standard deviation was 2.26%, which indicated that the accuracy of this method was relatively high. Compared with the traditional method, this method is more simple and rapid, and its accuracy can meet the needs of conventional oil content measurement, which is expected to realise the rapid on?site determination of oil content of large?scale oil?containing sludge in plants.