DIGITAL MEDIA PROFESSIONALS INC. | Making the Image Intelligent






2022国際ロボット展(iREX2022)デモ / Demonstrations at International Robot Exhibition 2022 (iREX2022)

2022国際ロボット展(iREX2022)デモ / Demonstrations at International Robot Exhibition 2022 (iREX2022)



図1: ドローンから撮影した映像 – 特定人物を解析中 / Figure1: Image taken from a drone – Specific person is being analyzed.
図2: ドローンから撮影した映像 – GAN技術を活用して様々な画像に加工(ゴッホ風等) / Figure2: Image take from a dron – Processed into various images using GAN technology(Van Gogh style, etc.)


図3: タフネスロボットアームのブース – ブースの上部にAIカメラを設置してデジタル安全柵を実現 / Figure3: Toughness Robot Arm Booth – A digital safety fence is realized by installing an AI camera at the top of the booth.
図4: デジタル安全柵内で人の手を検出したAIカメラ画像 – ロボットアームは自動的に停止 / Figure4: Image of the AI camera that detected a human hand inside the digital safety fence – The detection automatically stopped the robotic arm.

DMP demonstrated in the booth of Yamaha Motor, a business partner, at the “International Robot Exhibition 2022 (iREX2022)” (Tokyo Big Sight, March 9-12, 2022): real-time AI recognition using a UMS (drone) and a camera, a digital safety fence monitoring the perimeter of collaborative robots in real time, a static display of a monocular AI camera.

Real-time AI recognition demonstration with drone and camera
DMP’s proprietary AI image recognition technology using cameras analyzed human flow in real time from the drone’s camera, and alerted visitors for social distancing if the distance between visitors was within 50 cm for more than 15 minutes. It also recognized the gesture of visitors waving their hands and played a 10-second video of Yamaha Motor. The following two photos show images of the venue taken from a drone; figure 1 shows a specific person being analyzed, and figure 2 shows a processed image by using GAN (Generative Adversarial Network) technology. This AI recognition model can be applied not only to drones, but also to security cameras for analyzing gestures, headcount, and social distance.

Digital safety fence demonstration
DMP exhibited a real-time demonstration of a digital safety fence using an AI camera as a solution for monitoring the perimeter of collaborative robots. A camera that constantly monitors a certain area is installed above the Toughness Robot Arm that Yamaha Motor exhibited as a reference, and if, for example, a human hand enters the area, the robot arm shall automatically stop. Since a serious accident can occur when an operating collaborative robot bumps into a person, a digital safety fence that combines a camera and AI image recognition technology helps a robot perform its tasks safely. Conventionally, safety has been ensured by installing physical safety fences or utilizing sensors such as LiDAR, but by utilizing cameras, it can be ensured in an inexpensive way. DMP will actively promote the digital safety fences by introducing them to customers in combination with Cambrian products (handled by DMP), which are vision systems for robot arms, and by introducing them as optional products for robot arms such as Universal Robots to trading companies that handle robot systems.