The Line crossing detection parameters. The Administrator client
Parameters | A working name of a parameter | A parameter type | A default value | An acceptable range * | A parameter description |
|---|---|---|---|---|---|
Frames per second. | fps | float | 5.0 | [1, 30] | A certain number of frames per second a video is processed. |
Prediction confidence threshold. | confidence_threshold | float | 0.534488676 | (0, 1) | Detection score threshold applied for predicted objects. |
A minimum object size. | min_size | float | 0.01 | (0, 1] | A minimum relative size of a moving object to trigger detection. |
A maximum object size. | max_size | float | 0.9 | (0, 1] | A maximum relative size of a moving object to trigger detection. |
Motion area expansion | motion_margin | float | 0.5 | (0, +inf) | Additional context which would be taken around the movement area before passing it into detector. Denotes size relative to motion rect size. Thus 0 means no margin, 1 is to extend motion area two times etc. |
Event processing time. | aggregation_time | float | 1.0 | (0, +inf) | Time (in seconds) during which the best object snapshot is selected for recognition. Higher values increase the quality of recognition, but also increase the delay between object appearance and event geneation. |
Track lifetime without new events | track_lifetime | float | 5.0 | (0, +inf) | A track time of an object that is measured in seconds and doesn’t include new events. |
Time duration of sparse history | sparse_history_interval | float | 300.0 | (1., +inf) | Time interval (in seconds) that is used to keep sparse frames in history. Logically should be greater than dense_history_interval. |
Update frequency of sparse history | sparse_population_frequency | float | 60.0 | (1., +inf) | Frequency (in seconds) with wich we populate sparse history. Thus sparse_history_interval=300 and sparse_population_frequency=60 means that we keep 1 frame per minute for 5 minutes (5 frames). |
Immobility threshold | still_threshold | float | 0.5 | (0, 1) | Threshold used to eliminate false positive motion detections when an overlapped stationary object becomes visible again. The more the value the more previous frames the object should be present in to be considered as still object. |
Minimum similarity between the object and the track | similarity_threshold | float | 0.5 | (0, 1) | If the similarity of two face shots is below this value, they are logged as separate events. High values may result in event spam, and lower values may result in missing events. |
Person similarity threshold | person_similarity_threshold | float | 0.7 | (0, 1) | Person similarity threshold |
Visualization | prod_visualize | bool | True | {False, True} | Whether to visualize event frames. |
*In the Acceptable range column, square brackets “[” and/or “]” mean that the threshold values can be used as the default ones; parentheses “(” and/or “)” mean that the threshold values cannot be used as the default values.