No abstract

As a physiological disorder, chilling injury in kiwifruit may develop when the fruit are stored for long periods at a low storage temperature of 0–1°C. Presence of the disorder, inconsistent with marketing requirements for high-quality fruit, may lead to substantial financial and reputational losses. Thus, early detection or removal of chill-damaged fruit is desirable. This study demonstrates a novel dual-laser scanning system which has potential to be developed into a fast online system for the detection of chilling injury in Actinidia chinensis var. chinensis ‘Zesy002’ kiwifruit. The system consists of two laser modules at 730 and 880 nm wavelengths, a scanning mechanism and two detectors at partial (90°) and full (180°) light transmission. A sample of 231 kiwifruit was used to prove the concept, including 80 sound and 151 chill-damaged fruit of three different severity categories (slight, moderate and severe). A principal component analysis – back propagation neural network was used to classify fruit with 5-fold cross-validation. A comparison was made with standard visible-near infrared (Vis-NIR) interactance spectroscopy used to classify the same fruit using the same modelling algorithm. The dual-laser scanning system showed a slightly higher binary classification accuracy than the Vis-NIR spectroscopy, with an average accuracy of 95% for distinguishing sound and chill-damaged fruit. The classification error rate was 0% for severe damaged fruit. These experimental results demonstrate the potential of this dual-laser scanning system for the detection of chill-damaged fruit. The setup using only two wavelengths, its unique scanning operation and flexible system layout make it practical and attractive for future development for application on high-speed fruit graders.

In the field of seismic interpretation, univariate databased maps are commonly used by interpreters, especially for fault detection. In these maps, the contrast between target regions and the background is one of the main factors that affect the accuracy of interpretation. Since univariate data-based maps are not capable of providing a high-contrast representation, to overcome this issue, we turn them into multivariate data-based representations using color blending. We blend neighboring time sections or frames that are viewed in the time direction of migrated seismic volumes as if they corresponded to the red, green, and blue channels of a color image. Furthermore, to extract more reliable structural information, we apply color transformations. Experimental results show that the proposed method improves the accuracy of fault detection by limiting the average distance between detected fault lines and the ground truth into one pixel.