Kanzelh\"ohe Observatory (KSO) regularly performs high-cadence full-disk imaging of the solar chromosphere in the H$\alpha$ and CaIIK spectrallines as well as the solar photosphere in white-light. In the frame of ESA's Space Situational Awareness (SSA) programme, a new system for real-time H$\alpha$ data provision and automatic flare detection was developed at KSO. The data and events detected are published in near real-time at ESA's SSA Space Weather portal (http://swe.ssa.esa.int/web/guest/kso-federated). In this paper, we describe the H$\alpha$ instrument, the image recognition algorithms developed, the implementation into the KSO H$\alpha$ observing system and present the evaluation results of the real-time data provision and flare detection for a period of five months. The H$\alpha$ data provision worked in $99.96$% of the images, with a mean time lag between image recording and online provision of 4s. Within the given criteria for the automatic image recognition system (at least three H$\alpha$ images are needed for a positive detection), all flares with an area $\ge$50 micro-hemispheres and located within $60^\circ$ of the Sun's center that occurred during the KSO observing times were detected, in total a number of 87 events. The automatically determined flare importance and brightness classes were correct in $\sim$85%. The mean flare positions in heliographic longitude and latitude were correct within $\sim$1$^\circ$. The median of the absolute differences for the flare start times and peak times from the automatic detections in comparison to the official NOAA (and KSO) visual flare reports were 3 min (1 min).

Solar storms can have a major impact on the infrastructure of the earth. Some of the causing events are observable from ground in the H{\alpha} spectral line. In this paper we propose a new method for the simultaneous detection of flares and filaments in H{\alpha} image sequences. Therefore we perform several preprocessing steps to enhance and normalize the images. Based on the intensity values we segment the image by a variational approach. In a final postprecessing step we derive essential properties to classify the events and further demonstrate the performance by comparing our obtained results to the data annotated by an expert. The information produced by our method can be used for near real-time alerts and the statistical analysis of existing data by solar physicists.