A Comparative Study of Physically Accurate Synthetic Shadow Datasets in Agricultural Settings with Human Activity.

Shadow, a natural phenomenon resulting from the absence of light, plays a pivotal role in agriculture, particularly in processes such as photosynthesis in plants. Despite the availability of generic shadow datasets, many suffer from annotation errors and lack detailed representations of agricultural shadows with possible human activity inside, excluding those derived from satellite or drone views. In this paper, we present an evaluation of a synthetically generated top-down shadow segmentation dataset characterized by photorealistic rendering and accurate shadow masks. We aim to determine its efficacy compared to real-world datasets and assess how factors such as annotation quality and image domain influence neural network model training. To establish a baseline, we trained numerous baseline architectures and subsequently explored transfer learning using various freely available shadow datasets. We further evaluated the out-of-domain performance compared to the training set of other shadow datasets. Our findings suggest that AgroSegNet demonstrates competitive performance and is effective for transfer learning, particularly in domains similar to agriculture.

A synthetic shadow dataset of agricultural settings.

Shadow, a natural phenomenon resulting from the absence of direct lighting, finds diverse real-world applications beyond computer vision, such as studying its effect on photosynthesis in plants and on the reduction of solar energy harvesting through photovoltaic panels. This article presents a dataset comprising 50,000 pairs of photorealistic computer-rendered images along with their corresponding physics-based shadow masks, primarily focused on agricultural settings with human activity in the field. The images are generated by simulating a scene in 3D modeling software to produce a pair of top-down images, consisting of a regular image and an overexposed image achieved by adjusting lighting parameters. Specifically, the strength of the light source representing the sun is increased, and all indirect lighting, including global illumination and light bouncing, is disabled. The resulting overexposed image is later converted into a physically accurate shadow mask with minimal annotation errors through post-processing techniques. This dataset holds promise for future research, serving as a basis for transfer learning or as a benchmark for model evaluation in the realm of shadow-related applications such as shadow detection and removal.

Comparison of 2D and 3D convolutional neural networks in hyperspectral image analysis of fruits applied to orange bruise detection.

Recent advances in hyperspectral imaging (HSI) have demonstrated its ability to detect defects in fruit that may not be visible in RGB images. HSIs can be considered 3D images containing two spatial dimensions and one spectral dimension. Therefore, the first question that arises is how to process this type of information, either using 2D or 3D models. In this study, HSI in the 550-900 nm spectral range was used to detect bruising in oranges. Sixty samples of Thompson oranges were subjected to a mechanical bruising process, and HSIs were taken at different time intervals: before bruising, and 8 and 16 h after bruising. The samples were then classified using two convolutional neural network (CNN) models, a shallow 7-layer network (CNN-7) and a deep 18-layer network (CNN-18). In addition, two different input processing approaches are used: using 2D information from each band, and using the full 3D data from each HSI. The 3D models were the most accurate, with 94% correct classification for 3D-CNN-18, compared to 90% for 3D-CNN-7, and less than 83% for the 2D models. Our study suggests that 3D HSI may be a more effective technique for detecting fruit bruising, allowing the development of a fast, accurate, and nondestructive method for fruit sorting. PRACTICAL APPLICATION: Orange bruises can reduce the market value of food, which is why the food processing industry needs to carry out quality inspections. An effective way to perform this inspection is by using hyperspectral images that can be processed with 2D or 3D models, either with deep or shallow neural networks. The results of the comparison performed in this work can be useful for the development of more accurate and efficient bruise detection methods for fruit inspection.