Estimation of leaf area index using inclined smartphone camera

Measurements of leaf area index (LAI) are important for modeling microclimate in vegetation research. Among the instruments for measuring the LAI, smartphone cameras are becoming an attractive alternative to special LAI instruments. However, the narrow full field of view (FOV) of the common smartphones offer only an effective viewing zenith angle (VZA) of less than 35 degrees when the camera is pointing straight up. To overcome this limitation, we propose a method to estimate LAI from an inclined smartphone camera that can enlarge the range of the sensor's effective VZA. With the directional gap fractions extracted from the images taken by the inclined smartphone camera, a curve matching algorithm is used to iteratively search for the simulated G functions, i.e. functions of mean tilt angle (MTA) and VZA. The MTAs corresponding to the matched G functions are selected as ancillary parameters to help calculate the LAI. The proposed method is validated using data collected over crops and trees by a LAI-2200 instrument and a Huawei Honor 7 smartphone. The results reveal that an inclination angle of 30 degrees from zenith is superior to other angles of 0, 45 and 60 degrees. A good agreement between the LAI measurements from the proposed method and those from the LAI-2200 supports the accurate estimation of MTAs. The success of the MTA estimates and thus LAI measurements is attributed to the enlarged VZA ranging from 4 degrees to 60 degrees, and this VZA is comparable with that of the LAI-2200 instrument. The attraction of the proposed method is that it does not rely on the empiric al G value or MTA, providing an affordable alternative to traditional commercial instruments. Future efforts can be directed to automatically capture images when the smartphone is inclined to the desired angle.