Outdoor sports broadcast enhancement: AI lighting correction beyond football

LivePictureAI is real-time AI that balances bright sunlight and deep shadows in live sports. We built it for football, and football is where it has been tested. But the picture problem it tackles is not a football problem. It is a light problem, and it turns up wherever a camera points at an outdoor venue on a sunny day. If you are new to the topic, start with our overview ofshadows in live sports; this page looks at how the same issue shows up across other outdoor sports.

One problem, many venues: sun, shade and floodlights

A broadcast camera can only expose a frame one way at a time. When part of the shot sits in direct sunlight and part in deep shade, the operator has to choose: protect the highlights and let the shadows fall to near black, or open up for the shadows and let the sunlit areas blow out. Global adjustments to iris, gain or gamma move the whole picture together, so they cannot fix both regions within the same frame.

Stands, roofs, trees, grandstands and the shape of the venue decide where that shadow line falls, and the sun decides when it moves. Evening fixtures bring their own version of the problem, as fading daylight mixes with floodlights and leaves uneven pools of light across the field of play. The details change from venue to venue, but the result for the viewer is the same: players, balls and moments that disappear into the dark part of the frame.

LivePictureAI works on that problem directly. It processes the feed frame by frame and rebalances sunlit and shaded regions, so both stay readable at once. It runs as a processing layer on the camera feeds or on the programme output, needs no new cameras, and adds at most 100 ms of latency.

Rugby broadcast image enhancement

Rugby shares the same stadium geometry and wide shots as football. Large stands on one side throw a hard-edged shadow across the pitch during afternoon kick-offs, and the main camera position often looks straight across that line. Rucks and mauls can sit half in sun and half in shade, and a break down the shaded touchline can be hard to follow on a wide shot.

Because the camera positions and framing are so close to football, rugby is the most natural next step for rugby broadcast image enhancement. The same frame-by-frame rebalancing applies to the same kind of picture.

Cricket broadcast image enhancement

Cricket has a time problem more than a geometry problem. Play runs for long daytime sessions as the sun tracks right across the ground, so the shadow cast by stands and pavilions changes shape and position over the course of the day. A setting that works in the morning session may not work after lunch.

For cricket broadcast image enhancement, the attraction of a frame-by-frame approach is that it does not depend on a fixed correction. It responds to the light in each frame, whether that is a fielder in the deep standing in shade or the pitch itself split between bright sun and a lengthening shadow.

Tennis broadcast image enhancement

On an open court, the shadow of the stands travels across the playing surface as a match runs. Long matches can start in full sun and finish with one baseline in shade, which means a player can move from bright light into darkness within a single rally. The ball itself is small and fast, and losing it in the shaded part of the court is exactly the kind of moment viewers notice.

Tennis broadcast image enhancement has to keep both ends of the court readable without flattening the picture. Rebalancing sunlit and shaded regions separately, rather than shifting the whole frame, is designed to do that.

Baseball and American football

In baseball, the classic shot shows a shadow line between a sunlit infield and a shaded mound, often with the batter and catcher on one side and the pitcher on the other. That split runs through the most-watched part of the frame for most of the game.

American football has hard-edged stand shadows crossing the field at afternoon kick-off, much like football and rugby. Wide shots that follow a play down the field can cross from sun into shade and back several times in a few seconds.

Golf, cycling, motorsport, athletics, horse racing and alpine

The list goes on. If the sun reaches it, the problem is there. These sports add a twist: many of them are not played in a single stadium, so the light changes with location as well as with time.

• Golf: fairways lined with trees, greens half in shade and players walking in and out of shadow between shots.
• Cycling: riders moving through tree-lined roads, villages and open country, with light changing every few seconds.
• Motorsport: cars passing under bridges, grandstands and trees at speed, where a shaded section of track can swallow detail.
• Athletics: the stadium shadow across part of the track and infield, with events running simultaneously in sun and shade.
• Horse racing: long straights and bends that pass in and out of the shadow of stands and trees.
• Alpine skiing: bright snow next to deeply shaded sections of the course, one of the hardest contrasts a camera faces.

Each of these has its own production style, and we would not assume one result carries over without seeing the footage. But the underlying issue, bright and dark regions in the same frame, is the one LivePictureAI addresses.

What carries over from football

Our work so far has been on live football, where half the pitch in sun and half in shadow is a familiar sight. Several things from that work are not specific to the sport:

• The approach. The feed is processed frame by frame, rebalancing sunlit and shaded regions instead of applying one global adjustment.
• The integration. A processing layer on the camera feeds or on the programme output, working with SMPTE 2110, NDI, SDI, RTSP and file-based streams.
• The equipment. No new cameras. It works with the cameras and lenses already in place.
• The timing. A maximum of 100 ms of added latency, built to keep pace with live production.