The LASCO-C2 coronagraph on-board the SOHO solar observatory has been providing a continuous flow of coronal images for the past nine years. Synoptic maps for each Carrington rotation have been constructed from these images and offer a global view of the temporal evolution of the solar corona, particularly the occurrence of transient events such as the coronal mass ejections (CMEs), an important component of space weather activity. CMEs present distinct signatures on synoptic maps offering a novel approach to the problem of their statistical detection. We are presently testing several techniques of automatic detection based on their morphological properties. The basic procedure involves three steps: i) morphological characterization, ii) definition and application of adapted filters (optimal trade-off filters, Canny filter,...), iii) segmentation of the filtered synoptic maps. At this stage, the CMEs are detected. The efficiency of the detection of the various filters is estimated using the ROC curves. On-going studies include the classification of CMEs based on their physical properties, the determination of their velocities, and the question of their connection to the streamer belt.