One leaf carries enough information to name most trees: its overall shape (lobed, oval, needle), its margin (smooth, toothed, spiny), and its vein pattern (parallel, pinnate, palmate). Lay the leaf flat on a plain background (your palm works) and photograph it top-down, and Leafr reads those features and names the tree, with its look-alikes and how to tell them apart.
No leaves? Trees identify year-round: bark texture, branching pattern, buds, and fruit each work. Winter identification is slower but entirely possible: photograph the bark at chest height and the tree's overall silhouette.
The leaf photo that reads best
Flat, top-down, plain background, natural light: the vein network is the fingerprint, and it needs to be in focus. Include the stalk (petiole): its length and whether it's flattened separates species that leaves alone can't (poplars are the classic case). If the leaves are compound (many leaflets on one stalk), photograph the whole compound leaf, not one leaflet; mistaking a leaflet for a leaf is the most common tree-ID error there is.
Beyond the leaf: season-proof identification
Spring flowers, summer leaves, autumn fruit, winter bark: every season offers a feature set. Acorns, samaras (the helicopter seeds), cones, and berries are often more diagnostic than foliage. Bark identifies mature trees at a glance once you know the patterns: beech's smooth gray, birch's paper, pine's plates. Leafr accepts all of them; photograph whichever feature the season gives you.
Leafr AI: Tree Identifier
One leaf is enough. Leafr reads its shape and veins and tells you which tree you're under.