In recent years there has been an increased interest in ocean surveillance. The activity includes control and monitoring of illegal fisheries, manmade ocean pollution and illegal sea traffic surveillance, etc. The key problem is how to identify ships and ship-like objects accurately and in a timely manner. In this context, currently, many solutions have been proposed based on high resolution optical and radar remote sensing systems. Most often, these systems suffer from two major limitations viz., limited swath, thereby requiring multiple satellites to cover the region of interest and huge volumes of data being transmitted to ground, even though effective per-pixel information content is minimal. Another limitation is that the existing systems are either simulated on ground or built using the non-space qualified/Commercial Of-The-Shelf (COTS) components. This paper proposes an efficient on-board ship detection system/package connected with medium resolution wide swath optical camera. The methodology adopted has three major components, viz., onboard data processing for improving the radiometric fidelity, followed by a ship detection using modified Constant False Alarm Rate algorithm (CFAR) and a false alarm suppression module to mask false identifications. Finally, the package outputs only the locations of the ships, which is transmitted to the ground. The proposed system reduces the effective volume of data to be transmitted and processed on ground and also significantly cuts down the turnaround time for achieving the end objective. The system is built on radiation hardened Field Programmable Gate Array (FPGA) devices to meet the various engineering constraints such as real-time performance, limited onboard power, radiation hardness, handling of multiple custom interfaces etc. The system is tested with one of the medium resolution Multispectral Visual and Near Infra-Red (MX-VNIR) sensor having a spatial resolution of around 50 m and swath of around 500 Kms, which wo The key problem is how to identify ships and ship-like objects accurately and in a timely manner. In this context, currently, many solutions have been proposed based on high resolution optical and radar remote sensing systems. Most often, these systems suffer from two major limitations viz., limited swath, thereby requiring multiple satellites to cover the region of interest and huge volumes of data being transmitted to ground, even though effective per-pixel information content is minimal. Another limitation is that the existing systems are either simulated on ground or built using the non-space qualified/Commercial Of-The-Shelf (COTS) components. In [ 23 ] authors proposed Convolution Neural Network (CNN)-based ship detection using high resolution satellite imagery and in [ 24 ] the authors have used ground-based CCTV video images to detect and predict the ship behavior. Both these methods have focused on inshore ship detection and behavior analysis near the sea shore, however these solutions are impractical for deep sea monitoring and larger region surveillance. Active sensors such as SAR [ 2 , 6 , 25 , 26 , 27 , 28 ] -based imaging are more effective due to their all-time imaging and cloud penetration capability. Hence a system having coarser to medium resolution imaging capability, covering wider swath and higher revisit cycle would be an ideal solution for instantaneous detection of dynamic objects like a ship/vessel when the required region of interest spans a very large area. Due to technological advancements in the last decade, the technique of on-board processing of satellite data [ 21 , 29 , 30 , 31 ] and information extraction is becoming a research hotspot. Researchers are looking into the feasibility of enhanced hardware-software blending for meeting end application needs. A Look-Up-Table (LUT) is made by applying the inversion of LTC and normalizing it across all the CCDs within the detector array. Although the derived LUT is non-linear in general, but in practice a best fit linear curve is derived and applied for simplicity [ 21 , 33 , 34 ]. The NUC correction followed in this article is as follows: Let X i n be the i th raw pixel denoted in n-bit unsigned integer; and G i and S i be the corresponding gain and offset represented in signed fractional notation. Thus, the system is developed with 16-bit signed fractional data type representation with 1-bit for sign, 7-bits for the mantissa and 8-bits for the fractional part. Figure 7 Accuracy against Bit-Precision. 3.4. Configuration Based Execution The system is designed to be very generic and suitable for use with a wide variety of sensors. Sensor specific parameters are kept configurable to abstract from the implementation. The configurability accounts for many sensor dependent behavior as well as algorithm tuning during mission life. A push broom camera having up to 16K CCD elements with up to 12-bit pixel depth can be connected with the package through this interface. The current OSD package is implemented for an upcoming satellite having a push broom camera with six bands in MX-VNIR each having a detector array of 12K CCD elements with 12-bit pixel depth and 42m spatial resolution. ○ The package receives Ancillary (AUX) data pertaining to orbit and attitude sensor, UTC time and modes etc. from the on-board computer (OBC) through a 1553 IP interface. ○ Command and control for configuring and health monitoring of the package is done through a 1553 IP interface with OBC. If we further dwell on the fact the results in clear waters are much better than the above observation. In the future we propose to use multiple bands to improve the accuracy and also incorporate geocoordinate generation. We also intend to extend the same methodology for other objects/events such as the detection of forest fires, anomaly detections, etc., using similar sensors or other resource constrained missions like micro/nano satellites. Another line of work is towards extending