Singeing improves the visual quality of rind in pig slaughterhouses. In addition, the thermic inactivation allows the bacterial contamination of carcass surfaces to be reduced. However, the thermal image of pig carcasses shows temperature differences along the carcasses. The objective of this study was to develop a method to analyze a thermal image. After defining an imaging protocol adapted to pig slaughterhouses, image analysis was carried out with the free software: ImageJ©. Three methods were developed in order to preserve the spatial arrangement and time information for each measurement: (1) Method by line profile, (2) Method by mouse over, (3) Method by mouse over and line profile. Moreover, the method had to be simple and accurate. After a statistical comparison (mean, variance, distribution) between the three methods on 20 carcasses, the method by line profile was validated to analyze the characteristics of heat treatment in four pig slaughterhouses during process. This validated method of image acquisition and analysis was a tool used to measure the temperature variability on carcasses. The reference temperature was the temperature of carcasses after 90 minutes of continuous process. This method confirmed temperature differences on carcass surfaces (the top part of the carcass is warmer than the bottom part within a range of 2-7°C) and within a production day; the lowest temperatures were observed when the process started or restarted (after breaks), on average 4°C. These temperature variations along the working day were observed in every slaughterhouse studied. Bacterial contamination was evaluated on carcasses prior and following singeing for two heating settings. Results were similar to the literature (reduction of 3 Log10 CFU/cm² of the Aerobic colony counts). With this pragmatic method, companies could have access to a measurement tool not only for characterizing, checking, optimizing and qualifying the efficiency of singeing, but also for investigating all thermic treatments on meat in slaughterhouses.