Context. Accurate detections of frequent small-scale extreme ultraviolet (EUV) brightenings are essential to the investigation of the physical processes heating the corona.
Aims. We detected small-scale brightenings, termed campfires, using their morphological and intensity structures as observed in coronal EUV imaging observations for statistical analysis.
Methods. We applied a method based on Zernike moments and a support vector machine (SVM) classifier to automatically identify and track campfires observed by Solar Orbiter/Extreme Ultraviolet Imager (EUI) and Solar Dynamics Observatory (SDO)/Atmospheric Imaging Assembly (AIA).
Results. This method detected 8678 campfires (with length scales between 400 km and 4000 km) from a sequence of 50 High Resolution EUV telescope (HRIEUV) 174 E images. From 21 near co-temporal AIA images covering the same field of view as EUI, we found 1131 campfires, 58% of which were also detected in HRIEUV images. In contrast, about 16% of campfires recognized in HRIEUV were detected by AIA. We obtain a campfire birthrate of 2 E-10 16m 2s 1. About 40% of campfires show a duration longer than 5 s, having been observed in at least two HRIEUV images. We find that 27% of campfires were found in coronal bright points and the remaining 73% have occurred out of coronal bright points. We detected 23 EUI campfires with a duration greater than 245 s. We found that about 80% of campfires are formed at supergranular boundaries, and the features with the highest total intensities are generated at network junctions and intense H I Lyman-α emission regions observed by EUI/HRILya. The probability distribution functions for the total intensity, peak intensity, and projected area of campfires follow a power law behavior with absolute indices between 2 and 3. This self-similar behavior is a possible signature of self-organization, or even self-organized criticality, in the campfire formation process.