Open topic with navigation
Optimize Object Recognition Performance
The quality of the image or video that you send to Media Server can have a significant effect on the performance of object recognition.
Consider the following expectations for input images and video:
- The size of the object within the image or video frame is important. Object recognition works reliably (depending on other factors) if the object occupies a minimum area of 100x100 pixels. Some objects can be recognized down to 50x50 pixels, but Media Server does not usually recognize any object smaller than this. The size of the image or video is less important than the size of the object; however a large image might contain a large amount of clutter, which means that object recognition might take longer.
- Objects should not appear blurry.
- Objects can be recognized if they are partially obscured (set the
Occlusion configuration parameter), but this might reduce accuracy.
- Object recognition performs best when the image or video has even illumination, because dim lighting, dark shadows, or bright highlights can reduce accuracy.
- The image or video should not be affected by significant compression artifacts that can affect some formats (such as highly compressed JPEG images). HPE recommends that you do not save your image files as JPEG images with high levels of compression or transcode video that has been captured from a camera. If you are using a digital video recorder (DVR) to record the footage from a camera and are then sending the video to Media Server, ensure the DVR is saving the video at full resolution and is not reducing the bit rate.