Input Video Guidelines & Best Practices
This document aims to provide guidelines for Input Videos that can be used with Signzy technologies private limited. These guidelines will help you to ensure the best results for face search and recognition.
ThThere is no minimum requirement for image resolution, but the face resolution should be at least 50x50 pixels for images with a total resolution up to 1920x1080. Higher-resolution images require a larger minimum face size.
Please note that generating a high-resolution image from a low-resolution image does not produce the results needed for face search due to artifacts generated by the up-sampling of the image.
There are three measurements for camera angle - Pitch, Roll, and Yaw. For best results, we recommend:
- Pitch: Less than 30 degrees when the camera is facing down and less than 45 degrees when the camera is facing up.
- Roll: Less than 10 degrees in either direction.
- Yaw: Less than 45 degrees in either direction.
The face angle along an axis that is captured by the camera is a combination of both the camera angle facing the scene and the angle at hich the subject’s head is in the scene. For example, if the camera is 30 degrees facing down and the person has their head down a further 30 degrees, the actual face pitch as seen by the camera is 60 degrees. In this case, recognition would not be able to recognize the face. We recommend setting up cameras such that the camera angles are based on the assumption that people are generally looking into the camera with the overall pitch (a combination of face and camera) at 30 degrees or less.
The recommended minimum face resolution of 50x50 pixels should drive this camera setting. We recommend using the zoom setting of a camera so that the desired faces are at a resolution no less than 50X50 pixels.
The height of the camera plays an important role in capturing accurate face images. The recommended camera height should be such that it captures the face at a proper angle, and it should be based on the camera pitch angle.
When the camera is tilted down, the recommended pitch angle is less than 30 degrees, and when it is tilted up, the recommended pitch angle is less than 45 degrees. The height of the camera should be adjusted accordingly to capture the face at the recommended pitch angle.
It is also essential to note that the actual face angle along an axis captured by the camera is a combination of both the camera angle facing the scene and the angle at which the subject's head is in the scene. Therefore, it is crucial to set up cameras such that the camera angles are based on the assumption that people are generally looking into the camera with the overall pitch (a combination of face and camera) at 30 degrees or less.
Codec is a device that compresses and decompresses digital audio and video signals. The codec used for video recording plays a crucial role in the quality of the recorded video. The recommended codec for recording video is h.264 encoded.
H.264 is a widely used video compression standard that provides high-quality video at relatively low bitrates. It offers better compression than previous standards and is compatible with most devices and software applications.
Using the h.264 codec ensures that the recorded video is of high quality while maintaining low bitrates, which reduces the storage space required to store the video. This codec is widely supported and can be played on most devices and platforms, making it the preferred choice for video recording.
It is also recommended to use the h.264 codec for video transmission as it provides better video quality while consuming less bandwidth, making it suitable for real-time video streaming applications.
The recommended frame rate is 30 fps, and the recommended encoder bit rate is 3 Mbps. The frame rate should not be less than 5 fps, and the encoder bit rate should not be less than 1.5 Mbps.
If the encoder bitrate is a constraint, we recommend favoring a higher frame resolution over a higher frame rate for better face search results. This will ensure recognition gets the best quality frame within the allocated bit rate. However, there is a downside to this: due to the low frame rate, the camera will miss fast motion in a scene.
It is important to understand the trade-offs between these two parameters for a given set-up. For example, if the maximum possible bit rate is 1.5 MBPS, a camera can capture 1080p at 5 fps or 720p at 15 fps. The choice between the two is application dependent, as long as the recommended face resolution of 50x50 pixels is met.
The size of a video file is directly related to its bitrate and duration. The higher the bitrate and the longer the duration, the larger the video file size. Video size is a crucial factor to consider when it comes to API documentation, as it affects the storage space required to store the video and the time taken to transfer the video over a network.
It is recommended to keep the video size as small as possible without compromising the video quality. In general, a video size of up to 30 MB is considered acceptable.
Keeping the video size small has several advantages. Firstly, it reduces the storage space required to store the video. Secondly, it reduces the time taken to transfer the video over a network, which is essential for real-time video streaming applications.
However, reducing the video size too much can result in a loss of video quality. Therefore, it is important to strike a balance between video quality and video size, ensuring that the video quality is acceptable while keeping the video size as small as possible.
To achieve this, it is recommended to use appropriate video compression techniques to reduce the video size without compromising the video quality. Different video codecs and compression techniques can be used to achieve the best balance between video quality and size for different use cases. It is essential to test and optimize the video compression settings to achieve the best results.