An illumination invariant maize canopy structure parameters analysis method based on hemispherical photography

Abstract

Hemispherical photography (HP) has already proven to be a powerful indirect method for measuring various components of canopy structure. In this paper an illumination invariant multiple exposure images fusion and mapping method was proposed in order to squash negative impact of variant illumination. Firstly, a series of multiple exposure maize canopy hemispherical images was captured under natural light condition. Secondly, the multiple photographs fused into a single radiance map whose pixel truncated in shadowed and lighted parts of original images expended to higher range. We were able to determine the irradiance value at each pixel, the pixel values are proportional to the true irradiance values in the scene. The pixel values, exposure times, and irradiance values form a least squares problem. Finally, we also employed a histogram equalisation method to map irradiance values to RGB color space. The comparison results show that canopy gaps fraction of HP acquired at 14:00 and 17:00 with threshold value 180 has difference of 15.4% percent, and our method reduces the difference up to 2.8%. Results of regression analysis shows that our method have a high consistency with canopy structure parameter direct surveying method, the correlation coefficient between two methods hit 0.94. The line slope was 1.463, our method measurement values were lower than direct surveying method. Our method expands the HP canopy structure parameters acquire timing, provides an automatic monitoring solution.