Image-based comparison between a γ-ray scanner and a dual-sensor penetrometer technique for visual assessment of bale density distribution

There is an increasing desire to have an insight into the density distribution of silage bales because a low initial air content can significantly reduce the risk of temperature rise and the loss of sugar content. In addition, biomass moisture also plays a determinant role in the risk of temperature rise of bale and biomass degradation. For the visual assessment of the bale density, @c-ray scanner is regarded as an accurate and reliable technique but has potential risk of exposure to radiation. This study proposed a practical technique using a dual-sensor penetrometer for the simultaneous measurements of penetration resistance (PR) and moisture content (MC) in conjunction with image-based analysis. Four round bales made by two types of balers were prepared for the test. During a measurement process, both MC and PR data were collected following the signal of a penetrating depth transducer in-phase. Each bale included 72 penetration measurements that were transformed into an image referring to either PR or MC, respectively. With the reference to the two-dimensional (2D) results obtained from a @c-ray scanner, the dual-sensor images revealed the similar discrepancies of the baled density. Moreover, the proposed dual-sensor penetrometer method could discover the density discrepancies from the different sections of the measured bale so that the visualizations in three-dimensional (3D) space were also feasible. In contrast, the MC sensor seemed more reliable since the PR measurements were relatively susceptible to the interference from the penetrating friction between the shaft wall and material.