Tracker

or watch something

look

recognise things confidence 0.50 most objects 12 tracked points 130 small boxes 0.32 label size 23 lines box colour #ffffff

objects

box text: {class} {CLASS} {conf} {pct} {id} {x} {y} {w} {h} {v}

recognise things look for confidence 0.50 most boxes 12 numbers each 2 look every (ms) 100 box text where number from 1

subject

what moves by itself, with the camera taken out

ignore background 0.88 ignore camera sensitivity 0.50 memory 0.60 most subjects 6 smallest 0.12 show what it sees

find

what attracts a point

edges 0.50 motion 1.00 bright areas 0.00 dark areas 0.00 threshold 0.55 spacing 7 most points 130 detail 240

reach 6.0 linger 7 smoothing 0.30 first id 37000

boxes

how many 0.32 smallest 8 largest 110 style line 1.5 colour #ffffff colours opacity 0.85 big boxes 4 big box size 34 points to make one 7 big box style

labels

{id} {x} {y} {w} {h} {age} {v} {conf} {tag}

how many 1.00 text {tag} words font size 23 colour #ffffff where backing backing colour #000000

lines

shape hubs 3 lines each 6 longest 95 width 1.2 colour #ffffff opacity 0.75

trails

where things have been

length 0 behind width 2.0

patches

boxes filled with something

how many 0.05 kind colour #79dbe8

picture

brightness 0.00 contrast 1.00 saturation 1.00 hue 0 sharpen 0.00 solarize 0.00 solarize from 0.70 posterize 0 invert 0.00 palette palette amount 1.00 duotone dark #04060f duotone light #d8f7ff colour split 0.0 scanlines 0.00 grain 0.00 vignette 0.00 pixelate 1 glitch 0.00

overlay text

{rec} {clock} {date} {n} {regions} {counts} {list} {frame} {fps} {file}

text where size 26 colour #ffffff

output

playback speed 1.00 longest side 1920 recording Mbps 16