This study addresses the challenge of efficiently peeling pineapples, which have a distinct elliptical form, thick skin, and small eyes that are difficult to detect with conventional automated methods. This results in significant flesh waste. To improve the process, we developed an integrated system combining an enhanced BlendMask method, termed SAAF-BlendMask, and a Pose Correction Planning (PCP) method. SAAF-BlendMask improves the detection of small pineapple eyes, while PCP ensures accurate posture adjustment for precise path planning. The system uses 3D vision and deep learning technologies, achieving an average precision (AP) of 73.04% and a small object precision (APs) of 62.54% in eye detection, with a path planning success rate reaching 99%. The fully automated electromechanical system was tested on 110 real pineapples, demonstrating a reduction in flesh waste by 11.7% compared to traditional methods. This study highlights the potential of advanced machine vision and robotics in enhancing the efficiency and precision of food processing.
The study addresses the challenge of efficiently peeling pineapples, which have a distinct elliptical form, thick skin, and small eyes that are difficult to detect with conventional automated methods. This results in significant flesh waste. To improve the process, the researchers developed an integrated system combining an enhanced BlendMask method, termed SAAF-BlendMask, and a PCP method. SAAF-BlendMask improves the detection of small pineapple eyes, while PCP ensures accurate posture adjustment for precise path planning. The system uses 3D vision and deep learning technologies, achieving an average precision (AP) of 73.04% and a small object precision (APs) of 62.54% in eye detection, with a path planning success rate reaching 99%. The fully automated electromechanical system was tested on 110 real pineapples, demonstrating a reduction in flesh waste by 11.7% compared to traditional methods. This study highlights the potential of advanced machine vision and robotics in enhancing the efficiency and precision of food processing.
Three-dimensional instance segmentation has developed rapidly in the field of autonomous driving, with recent research primarily focusing on outdoor large-scale scenes. However, in robot operation environments characterized by target stacking, occlusion, and clutter, such studies often suffer from instance confusion, information loss of small and occluded targets, leading to insufficient segmentation accuracy. This makes them unsuitable for 3-D instance segmentation tasks in robotic manipulation scenarios. To address the challenges, we innovatively propose an adaptive fine-grained instance segmentation network (FGISN) based on a monocular depth sensor system, enabling precise object segmentation in complex scenes. The core of FGISN mainly comprises two components: category-space hybrid enhanced upsampler (CSHEU) and instance scale-aware attention fusion (SAAF) module. CSHEU employs a hybrid interpolation strategy that fuses semantic and spatial feature information to suppress noise interference from neighboring objects and resolve information confusion between different instances. SAAF leverages a cross-scale attention mechanism and a multiscale fusion method to achieve maximum sampling of information for small and occluded objects. Extensive experiments conducted on public datasets fully confirm the effectiveness of the proposed method.