Common problems of AGV vision system

05/15/202017:22:46 Comments 516

AGV robot vision can improve automated settings. The integrated robot solution can provide robot vision advantages quickly and easily without programming skills. However, even if the technology is improved, vision is also a tricky problem for AGV robot vision system technology.
Common problems of AGV vision system
The most common function of the AGV vision system is to detect the position and direction of known objects. The following factors affect the robot's vision system in the environment.

1. Lighting

If anyone has taken photos in low light, they will know that lighting is very important. If the lighting settings are not good, it will ruin this photo. The imaging sensor is not as adaptable as the human eye. If the type of lighting is wrong, then the visual sensor will not be able to reliably detect the object.

2. Deformation or articulation

The sphere is a simple object detected with computer vision settings. You may just detect its circular outline, or maybe use a template matching algorithm. However, if the sphere is squashed, it will deform and the same method will no longer work. It can cause considerable problems in robot vision technology.

The hinge resembles the deformation caused by a movable joint. For example, when your elbow bends your arm, the shape of the arm changes. Each link (bone) maintains the same shape, but the outline is deformed. The AGV vision system algorithm uses shape contours, so clarity makes object recognition more difficult.

3. Position and direction

The most common function of the AGV vision system is to detect the position and orientation of known objects. Therefore, most integrated vision solutions usually overcome the challenges faced by both. As long as the object can be viewed in the camera image, the location of the detected object is usually straightforward. However, not all directions are equal. Although it is simple enough to detect an object rotating along one axis, it is more complicated to detect when the object rotates in 3D.

4. Image background

The background of the image has a great influence on the ease of object detection. For example, put an object on a piece of paper, and then print the same object image on the paper. In this case, the robot vision system may not be able to confirm which is true.

The perfect background is blank and provides good contrast to the detected objects. Its exact properties will depend on the visual detection algorithm being used. If an edge detector is used, the background should not contain sharp lines. The color and brightness of the background should also be different from the color and brightness of the object.

5. Image occlusion

The image is blocked. In the first four questions, the entire object appears in the camera. The occlusion is different because some objects are missing. The visual system obviously cannot detect parts that do not exist in the image. There are many things that can cause occlusion, including: other objects, parts of the robot, or bad positions of the camera. Methods to overcome occlusion usually involve matching the visible part of the object to its known model and assuming that the hidden part of the object exists.

6. Size ratio

In some cases, the human eye is easily deceived by differences in scale. The AGV vision system may also be confused by them. Imagine you have two identical objects, but one is bigger than the other. Imagine that you are using a fixed 2D vision setting, and the size of the object determines its distance from the robot. If you train the system to recognize smaller objects, you will falsely detect that the two objects are the same and that the larger object is closer to the camera.

Another issue of scale, perhaps less obvious, is the issue of pixel values. If the robot camera is placed far away, the objects in the image will be represented by fewer pixels. When more pixels represent the object, the image processing algorithm will work better, with some exceptions.

7. Setting of camera position

An incorrect camera position may cause any problems that have occurred before, so it is important to use it correctly. Try to place the camera in a well-lit area so that you can see the object as clearly as possible without distortion, as close to the object as possible without causing obstruction. There should be no disturbing background or other objects between the camera and the viewing surface.

8. Object movement

When the object moves, it will cause problems with the visual settings on the computer, especially when the image is blurred. For example, this can happen to objects on fast moving conveyor belts. The digital imaging sensor captures the image in a short time, but does not capture the entire image in an instant.

If an object moves too fast during the capture process, this will cause the image to be blurred. Our eyes may not notice the blur in the video, but the algorithm will. The AGV vision system works best when there are clear still images.

Comment

You must beto post a comment.