Conventional control systems for prosthetic hands use myoelectric signals as an interface, but it is impossible to realize complex and flexible human hand movements with only myoelectric signals. A promising control scheme for prosthetic hands uses computer vision to assist in grasping objects. It features an imaging sensor, and the control system is capable of recognizing an object placed in the environment. Then, a gripping pattern can be selected from some predefined candidates according the recognized object. However, previous studies assumed that only one object exists in the environment. If there are multiple target objects in the environment, the hand could become confused in attempting to find the target object. This study addresses this problem and proposes a method to determine the target object from multiple objects. The proposed method is able to determine the target object by estimating the positional relationship between the artificial hand and the objects, as well as the motion of the hand. To verify the validity and effectiveness, we implemented the proposed method in a vision-based prosthetic hands control system and conducted pick-and-place experiments. The experiments confirm that the proposed method can accurately estimate the target object in accordance with the user’s intention.

Modern production lines for molded plastic parts often have automated inspection systems to detect defective parts reliably and efficiently. However, these conventional inspection systems have low flexibility and versatility, leading to difficulties when dealing with complicated requests such as when small quantities of many different parts are manufactured on the same production line. The proposed system can be implemented quickly using low-cost off-the-shelf components and does not require accurate alignment of production parts, reducing the need for manual inspections and increasing work efficiency when handling complex workloads. The inspection algorithm combines Higher-order Local Auto Correlation (HLAC) features with One-class Support Vector Machine (One-class SVM) and Principal Component Analysis (PCA) to extract, transform, and classify the differential feature vector between conforming and nonconforming plastic parts. To verify its validity and effectiveness, we compared defect detection accuracy and speed between the developed inspection system and manual inspection experimentally. Extremely high accuracy (Recall = 0.93, Specificity = 1.00) and speed (10 inspections in 30[sec]) was obtained with 7 types (1 conforming type, 6 nonconforming types) of sample parts (30 samples each). We demonstrated a 400 % increase in speed can be gained relative to manual inspection.

The estimation of leaf area is an important measure for understanding the growth, development, and productivity of tomato plants. In this study, we focused on the leaf area of a potted tomato plant and proposed methods, namely, NP, D², and D³, for estimating its leaf area. In the NP method, we used multiple tomato plant images from different viewing angles to reduce the estimation error of the leaf area, whereas in the D² and D³ methods, we further compensated for the perspective effects. The performances of the proposed methods were experimentally assessed using 40 “Momotaro Peace” tomato plants. The experimental results confirmed that the NP method had a smaller mean absolute percentage error (MAPE) on the test set than the conventional estimation method that uses a single tomato plant image. Likewise, the D² and D³ methods had a smaller MAPE on the test set than the conventional method that did not compensate for perspective effects.

In this paper, we propose a novel control scheme for a vision-based prosthetic hand. To realize complex and flexible human-like hand movements, the proposed method fuses bimodal information. Combining information from surface EMG signals with object information from a vision sensor, the system can select an appropriate hand motion. The training/recognition using both sEMG signals and object images can be performed with a single deep neural network in an end-to-end manner. The bimodal sensor information enables the system to recognize the operator's intended motion with higher accuracy than that of the conventional method using only sEMG signals. In addition, the generalization ability of the network is improved, so motion recognition robustness is enhanced against abnormal data that include partly noisy or missing samples. To verify the validity of the proposed approach, we prepared a dataset that contains the sEMG signals and the object images for 10 types of grasping motions. Three kinds of experiments were conducted: comparison of the proposed method with the conventional method, examination of the recognition robustness against partly noisy or missing samples, and challenges to recognize hand motions based on raw sEMG signals. The results revealed that the proposed bimodal network achieved considerably high recognition performance.

Model-based reinforcement learning is a powerful paradigm for learning tasks in robotics. However, real world learning tasks often involve complex patterns of missing data, and model-based reinforcement learning cannot handle missing data directly. To overcome this problem, in this paper, we focus on M-PGPE(GP) proposed by Mori et al. as a model-based reinforcement learning, and propose an extension of M-PGPE(GP) to handle missing data, which we call MM-PGPE(GP). The performance of the proposed MM-PGPE(GP) is assessed in two experiments with mountain car task. These experiments highlight the MM-PGPE(GP) produces higher average return and outperforms the conventional M-PGPE(GP) with simple linear interpolation.

Appearances of products are important to companies as they reflect the quality of their manufacture to customers. Nowadays, visual inspection is conducted by human inspectors. This research attempts to automate this process using Convolutional AutoEncoders (CAE). Our models were trained using images of non-defective parts. Previous research on autoencoders has reported that the accuracy of image regeneration can be improved by adding noise to the training dataset, but no extensive analyse of the noise factor has been done. Therefore, our method compares the effects of two different noise patterns on the models efficiency: Gaussian noise and noise made of a known structure. The test datasets were comprised of “defective” parts. Over the experiments, it has mostly been observed that the precision of the CAE sharpened when using noisy data during the training phases. The best results were obtained with structural noise, made of defined shapes randomly corrupting training data. Furthermore, the models were able to process test data that had slightly different positions and rotations compared to the ones found in the training dataset. However, shortcomings appeared when “regular” spots (in the training data) and “defective” spots (in the test data) partially, or totally, overlapped.