An Image‐Processing Method to Detect Planetary Transits: The “Gauging” Filter

A new method is proposed for signal denoising and transit detection in stellar photometric light curves. The plane of a light curve can be considered as a two-dimensional image that splits into two parts: the upper and lower part of the plot itself. The two resulting images are then analyzed with an image-processing method that enables us to develop new algorithms for filtering and detection. The resulting denoising algorithm is more efficient than standard filters, such as the convolution filter, and can be applied recursively without changing the signals of interest. The new detection algorithm has a performance similar to that of the matched filter, but also exhibits greater "sensitivity" at a given confidence level. These new filtering and detection algorithms can operate on selected parts of a light curve, which is important for processing data flows or analyzing interrupted data sequences. They have been developed and tested on simulated light curves created using a physical model of the COROT (Convection, Rotation, and Planetary Transits) instrument being developed for the mission. The algorithms provide interesting perspectives on how the detection of weak transit events might be improved.