Sensor data from image generating sensors require large transmission and storage capacity. Especially sensors with a high spatial, spectral or temporal resolution will create huge amounts of data. To handle this compression is needed. For applications in the defence area other demands are required from compression algorithms then what they are usually developed for. The compression methods of today are made for grayscale or color within the visual spectrum. Distortion given by the compression is optimized to give a minimum of artifacts for the human visual system. For defence applications other types of sensors are of interest and the distortion must be adapted for automatic image and signal processing methods such as detection. Within the project Sensor integrated data compression different types of coding methods have been tested on IR, SAR and hyperspectral sensors. After encoding and decoding the sensor data the results given by different detection algorithms have been investigated. The compression distortion is not only a disadvantage to the detection performance. In some cases the compression distortion can have the effect of a pre-filtering which enhances properties that improves the performance of the detection algorithm. But, in most cases the compression distortion is a disadvantage to the detection algorithm. Since the final result will depend on the parameter settings on both the compression algorithm and the detection algorithm a good performance measure is needed. We have used two different measures to be able to validate the performance on detection from compressed sensor data. The project has for the most part been working on real sensor data but we have also implemented a way to use compression in a sensor and signal processing simulation lab.