There are several obstacles and challenges that can occur when incorporating machine vision systems in automated operations. Here are a few common issues with 3D image processing and how Liberty Reach has meticulously designed systems to overcome those challenges:
Occlusion: Occlusion occurs when one object in a scene blocks the view of another object. This blocking makes it difficult to fully capture and process the 3D information.
Solution: Liberty Reach provides flexible sensor positioning and the ability to evaluate the cropped region of the image.
Lighting: Lighting conditions can cause significant variations in the appearance of objects in a 3D scene, making it difficult to accurately process and interpret the 3D information without system specific, controlled lighting.
Solution: The 3D near IR depth VSx sensors capture the data in the scene. VSx sensors operate outside of the visible light spectrum so they are not influenced by ambient light conditions.
Noise: Noise in 3D images can be caused by a variety of factors, including the imaging sensor, lighting conditions, and the environment (i.e. smoke, water vapor, or other abnormalities). This noise can make it difficult to accurately process and interpret the 3D information.
Solution: Our systems utilize proprietary noise removal algorithms to reduce noise in the image, allowing for offset calculations in harsh factory environments without causing faults and downtime.
Geometry: Complex and irregularly shaped objects can be difficult to fully capture and process using 3D imaging techniques, as they may not conform to typical assumptions and models used in 3D image processing.
Solution: Rigid objects can be measured using any surface of the part, without the need to map the whole part surface. Flexible sensor positioning and standoff distance also allows the system to work within customer restrictions to achieve a valid sensor view.
Time: In dynamic scenes, the movement of objects or the changes in lighting conditions can make it difficult to accurately capture and process the 3D information in real-time.
Solution: Liberty Reach guidance systems utilize a process called trigger mode to synchronize devices. Learn more about it here.
Ambiguity: 3D images can often be ambiguous, making it difficult to interpret and understand the 3D information. This can be due to the complexity of the scene, the viewpoint from which the image was captured, or the limitations of the imaging sensor.
Solution: Our systems only evaluate the cropped region of the image and reject the offset if the surface geometry does not match the trained model. The software also overlays the color image on the depth image so operators can easily interpret the scene in a way that they are accustomed to.
Limited data: 3D image processing often requires large amounts of data to accurately capture and process the 3D scene. Limited data can make it difficult to achieve accurate results.
Solution: The variable sensor standoff distances allows users to adjust the field of view and gain more or less data from the image of the part.
These are some examples of issues that can arise in 3D image processing. Liberty Reach mitigates these issues using a combination of knowledge, hardware, setup techniques, and advanced algorithms. Our products use different combinations of hardware and algorithms to protect the integrity of the scene, reduce noise, remove ambiguity, and handle the demands of data management quickly and efficiently. For more system or application information, sign up for a free demonstration.
To stay current on all things Liberty Reach, follow us on LinkedIn.
About Liberty Reach
Liberty Reach provides 3D volumetric vision guidance systems for robot applications. Our systems improve manufacturing, material/part handling, and logistics operations performed in industrial environments. We provide value by decreasing cycle time, increasing capacity and throughput, and improving automated operations. Furthermore, Liberty Reach 3D Vision Guidance systems increase repeatability and decrease faults and costly downtime. As the first company to deploy volumetric sensor technology on a high-precision manufacturing line, we are the experts in volumetric vision guidance solutions.