The Linux Foundation Projects
Skip to main content
All Posts By


Klepsydra AI – Cloud detection onboard from space

By Blog, Featured

Snapshot: Klepsydra OBPMark-ML Cloud Detection Demo in Progress

Klepsydra AI – Cloud detection onboard from space with a custom Linux distribution built by the Yocto Project


What is Cloud detection?

Cloud detection is a crucial process in Earth Observation used to identify and mask clouds in satellite imagery. This is necessary because clouds can obstruct the view of the Earth’s surface, making it difficult to accurately interpret and analyse the data.

Cloud detection algorithms typically use a combination of spectral and spatial information to differentiate between clouds and other features in the imagery. For instance, they may utilize information from various wavelengths of light to distinguish between clouds and land or water surfaces. They may also use contextual information, such as the size and shape of features in the image, to aid in cloud identification.

Once clouds are detected, they can be masked or removed from the image so that the underlying land or water surface can be analysed. This is important for a wide range of applications, including land use and land cover mapping, crop monitoring, and climate studies.

Cloud detection is also utilised in real-time applications such as weather forecasting and disaster management, where monitoring cloud cover and its changes over time is crucial. In these applications, cloud detection algorithms can track the movement and formation of clouds, providing valuable information for predicting weather patterns and identifying areas that may be impacted by natural disasters.

Klepsydra AI Excels Beyond Tensorflow Lite in Performance

Cloud detection onboard

Performing cloud detection onboard Earth Observation satellite offers several benefits over performing cloud detection on the ground:

  1. Faster response time: Cloud detection onboard the satellite enables near real-time detection and removal of clouds, which is particularly useful for time-critical applications such as weather forecasting and disaster response.
  2. Reduced data transmission: Transmitting large amounts of satellite imagery data to the ground can be expensive and time-consuming. By performing cloud detection onboard, only the useful data (i.e. data without clouds) needs to be transmitted to the ground, reducing data transmission costs.
  3. Improved data quality: Cloud detection onboard the satellite can result in improved data quality because the detection algorithms can take into account the unique characteristics of the satellite’s sensors and the viewing geometry. This can result in more accurate and reliable cloud detection.
  4. Increased availability of cloud-free data: By performing cloud detection onboard, the satellite can provide a higher percentage of cloud-free data, which is particularly important for applications such as land use and land cover mapping, crop monitoring, and climate studies.
  5. Improved efficiency of downstream processing: Cloud detection onboard the satellite can improve the efficiency of downstream processing by reducing the amount of data that needs to be processed on the ground. This can lead to faster and more accurate analysis of the data.

In Collaboration with ESA, Barcelona Supercomputing Center developed OBPMark-ML’s Cloud Detection Algorithm

KATESU project

The current commercial version of Klepsydra AI has successfully passed validation in an ESA activity called KATESU for Teledyne e2v’s LS1046 and Xilinx ZedBoard onboard computers, achieving outstanding performance results. During this activity, two DNN algorithms provided by ESA, CME and OBPMark-ML, were tested.

Onboard cloud detection is important to filter images that are sent to ground.

Klepsydra on a custom Linux distribution built by the Yocto Project

The Yocto Project is an umbrella organization for a number of open-source technologies which simplify the process of building and customizing Linux-based operating systems for embedded devices. It provides a flexible and scalable infrastructure, enabling developers to create highly optimized and tailored Linux distributions for their specific embedded systems.

In Klepsydra, the Yocto Project plays a crucial role in our workflow, particularly when it comes to generating Linux images for our LS1046 based computer. To begin, we set the necessary Yocto Project build tools up and configure the build system to specifically target our desired hardware platform. This ensures that the resulting Linux image is optimized and compatible with our LS1046 device.

To tailor the Linux kernel and other system components to our specific requirements, we create custom Yocto Project recipes. These recipes allow us to incorporate the necessary changes and optimizations, ensuring that the resulting Linux image is finely tuned to meet our needs. Additionally, we introduce the meta-virtualization layer to our Yocto Project setup, which enables us to include Docker in the final root filesystem of the generated image.

Once the build has been configured and our customizations have been applied, the next step is to generate the Linux image for our target platform. This involves compiling and packaging all the requiredcomponents, including the kernel, device drivers, libraries, and applications, into a deployable image file. This image file can then be flashed onto an SD card, effectively preparing it for booting Linux on our LS1046 device.

With the Linux image successfully booted on the target platform, we proceed to create the necessary Docker image directly on the LS1046 device. Leveraging the Docker capabilities provided by the meta-virtualization layer, we prepare a Docker image that encapsulates the specific software, dependencies, and configurations required for our testing purposes.

Once the Docker image is prepared, we launch a container from it on the target device. This container serves as a controlled environment where we can perform various tests and evaluations on the LS1046 device. By executing tests within the container, we can isolate and evaluate specific functionalities or scenarios, ensuring the reliability and performance of our software on the target platform.

Demo online

The demo showcases the Cloud Detection DNN model executed on three identical computers, each with a different optimisation. The first computer runs Klepsydra optimised for latency (, the second uses TensorFlow Lite, and the third uses Klepsydra optimised for CPU (kpsr.cpu).

Klepsydra AI demonstrates remarkable elasticity and high-performance capabilities. The configuration can process up to two times more images per second than TensorFlow Lite, while kpsr.cpu processes the same number of images as TensorFlow Lite but with fewer CPU resources. These improvements are evident in both the Intel and ARM versions of the demo.

In summary, Klepsydra AI provides customers with a unique capability to adapt to their specific needs, whether it be latency, CPU, RAM, or throughput. This feature makes Klepsydra AI highly suitable for onboard AI applications such as Earth Observation onboard data processing and compression, vision-based navigation for in-orbit servicing, and lunar landing.


This demo was prepared as part of ESA’s KATESU project to evaluate Klepsydra AI for Space use. For further information on this project, please refer to

The OBPMark-ML DNN was provided to Klepsydra by courtesy of ESA. This algorithm is part of ESA’s OBPMark framework ( For further information on this framework, please contact

PR: Exein, 4Y LTS, YPDD 2023

By Blog


Yocto Project Welcomes Exein as a Platinum Member, Announces Extended LTS Release Plan and One-Day Technical Summit

Yocto Project Invests in the Long Term, Continues Growth and Commits to Security with new Platinum Member.

SAN FRANCISCO – June 26, 2023 – The Yocto Project, an open source collaborative initiative helping developers create custom Linux-based systems, has evolved significantly over the last 12 years to meet the requirements of its community. The project continues to lead in build system technology with field advances in build reproducibility, software license management, SBOM compliance and binary artifact reuse. In an effort to support the community, The Yocto Project announced the first Long Term Support (LTS) release in October 2020. Today, we are delighted to announce that we are expanding the LTS release and extending the lifecycle from 2 to 4 years as standard.

The continued growth of the Yocto Project coincides with the welcomed addition of Exein as a Platinum Member, joining AMD/Xilinx, Arm, AWS, BMW Group, Cisco, Comcast, Intel, Meta and WindRiver. As a Member, Exein brings its embedded security expertise across billions of devices to the core of the Yocto Project. 

“Long Term Support (LTS) is one of our most asked about features, it is great the project is now able to commit to 4 years as standard for all our LTS releases,” said Richard Purdie, Yocto Project Lead Architect and Linux Foundation Fellow. “New members like Exein bring both specialist knowledge and funding, enabling us to do this and more. Exein’s involvement will truly bolster the security capabilities of the Yocto Project. The ability to offer enhanced embedded security is a major advancement in our pursuit of safer, more resilient systems.”

“The Yocto Project has been at the forefront of OS technologies for over a decade,” said Andrew Wafaa, Yocto Project Chairperson. “The adaptability and variety of the tooling provided are clearly making a difference to the community, we are delighted to welcome Exein as a member as their knowledge and experience in providing secure Yocto Project based builds to customers will enable us to adapt to the modern landscape being set by the US Digital Strategy and the EU Cyber Resilience Act” 

“We’re extremely excited to become a Platinum Partner of the Yocto Project,” said Gianni Cuozzo, founder and CEO of Exein. “The Yocto Project is the most important project in the embedded Linux space, powering billions of devices every year. We take great pride in contributing our extensive knowledge and expertise in embedded security to foster a future that is both enhanced and secure for Yocto-powered devices. We are dedicated to supporting the growth of the Yocto Project as a whole, aiming to improve its support for modern languages like Rust, and assist developers and OEMs in aligning with the goals outlined in the EU Cyber Resilience Act.” 

The Yocto Project is also excited to be hosting Yocto Project Dev Day on June 26 alongside the Embedded Open Source Summit in Prague, Czech Republic. Back for the first time since 2019, Yocto Project Dev Day brings together developers from across the Yocto ecosystem to participate in a variety of community sessions, presentations, and tutorials. For more information about the Yocto Project, visit


Media Contact

Noah Lehman

The Linux Foundation