3/13/2024 Embedded Vision Expert knowledge embedded world

embedded award 2024: Embedded vision nominees

The data volumes and data rates are very high for embedded vision systems. Consequently, the demands on the systems, their components and their connections are much higher than for classic sensor systems. As a result, embedded vision is driving the entire industry. The nominees in this category are also among the drivers of the industry.

Camera in the embedded vision area at the embedded world Exhibition&Conference

Embedded vision is driving the entire industry

A 3D Stereo Camera with Edge-AI, an Event Based Vision Sensor and AI Vision Processors

Hailo-15 AI Vision Processors

Exhibitor: Hailo
Hall/Booth: 1/1-126

Nowadays, AI is applied to practically all application domains. A significant challenge faced by smart cameras today is insufficient computing power to generate superior image quality and perform video analytics. Users are forced to make a trade-off between achieving excellent video quality or implementing effective analytics with multiple data points.

With the introduction of the Hailo-15 AI vision processors, this compromise is no longer necessary. Now, you can enjoy both AI-powered video quality enhancement and conduct multiple analytics with high accuracy at high resolution. With up to 20 AI TOPS and a dedicated neural network core, the Hailo-15 family of SoCs allows deployment of multiple AI models in parallel, running at high FPS.

This enables unparalleled AI analytics with fast and accurate detection of more objects per frame, and higher frame rate. Hailo-15 also offers premium 4K60 image quality with cutting-edge vision pipelines enabling low-light noise reduction, HDR and other video enhancement features. The result is an ultra-high-quality AI-based video analytics solution that keeps people and environments safer, while ensuring their privacy and allows organizations to operate more efficiently at a lower cost.

The Hailo-15 family of high-performance AI vision processors integrate directly into intelligent cameras, bringing unprecedented AI performance to cameras and enabling advanced video analytics and video enhancement for high image quality. Hailo-15 represents a significant step forward in making AI at the edge more scalable and affordable.

With its superior AI capacity, Hailo-15-empowered cameras can carry out significantly more video analytics, running several AI tasks in parallel, including faster detection at high resolution to enable the identification of smaller and more distant objects with higher accuracy and fewer false alarms. The stronger AI capacity can also be used for video enhancement and better video quality with a variety of features including low-light denoising, HDR and 3A (auto focus, auto white balance, auto exposure), distortion correction, motion stabilization, and digital zoom.

Hailo-15 supports many exciting and relevant applications including increased accuracy in industrial cameras performing quality inspection, pick & place or product identification, increasing safety in smart cities by quickly and accurately identifying and alerting to incidents and emergencies. It aids transportation authorities in prompt incident detection and response, addressing issues like lost children and mishandled luggage. In retail, Hailo-15 improves customer satisfaction with accurate insights and assists in stock level maintenance.

Would you like to delve deeper into the topic?
At embedded world Exhibition&Conference 2025 from March 11 to 13, 2025,
you will have the opportunity to exchange ideas with industry experts.

To the TicketShop

LIPSedge™ S-Series High-Resolution 3D Stereo Camera with Edge-AI Enabled

Exhibitor: LIPS Corporation
Hall/Booth: 2/2-658

The absence of high-resolution 2D and 3D image capabilities, along with the lack of edge computing functionality, poses significant challenges in diverse industrial applications. The deficiency in high-resolution imaging for both 2D and 3D dimensions compromises data accuracy, impeding critical tasks like quality control, object recognition, and inspection processes.

In sectors such as robotics and Autonomous Mobile Robots (AMR), the inability to simultaneously acquire detailed 2D and 3D data leads to diminished precision in object recognition and manipulation. Furthermore, the lack of edge computing capability in the camera device gives rise to latency issues and bandwidth constraints, necessitating the transmission of camera data to remote servers for processing.

This reliance on centralized processing not only raises privacy and security concerns but also restricts functionality in offline environments, fostering increased dependence on expensive cloud services. The cumulative impact of these limitations reverberates across various industries, including manufacturing, healthcare, and virtual and augmented reality, detrimentally affecting the efficiency, accuracy, and overall capabilities of industrial processes. Addressing these challenges requires the integration of high-resolution 2D and 3D camera capabilities with edge computing, offering tangible benefits such as improved performance, reduced latency, enhanced privacy, and increased flexibility in deployment.

The LIPSedge™ S-Series emerges as an innovative solution, strategically designed to address challenges in industrial automation. The device seamlessly integrates the Ambarella CV2 CVflow® SoC, delivering high-performance edge AI and image processing capabilities, including the essential deep neural network (DNN) processing required for intricate tasks.

As a next-gen 3D camera series, it goes beyond conventional capabilities by harmonizing high-resolution 2D and 3D image processing within a single camera, with roadmap achieving up to 4K RGB-D resolution. Specifically engineered for industries demanding a convergence of 2D imaging and 3D machine vision capabilities, the LIPSedge™ S-Series redefines possibilities. Its compact form factor is coupled with robust industrial capabilities, offering a versatile solution for a range of applications.

This device facilitates GigE/PoE data and power transmission across distances of up to 100 meters, housed within an IP67-rated rugged enclosure that ensures resilience against dust and water. The incorporation of a built-in IMU and heatsink enhances its adaptability for various applications, including but not limited to robotic pick & place, depalletizing, visual inspection, and Autonomous Mobile Robots (AMR). In essence, the LIPSedge™ S-Series not only meets the demands of industrial automation but surpasses expectations, presenting a sophisticated solution that adeptly tackles the identified limitations in the industry.