Rapid and accurate counting and recognition of flying insects are of great importance, especially for pest control. Traditional manual identification and counting of flying insects is labor intensive ...and inefficient. In this study, a vision-based counting and classification system for flying insects is designed and implemented. The system is constructed as follows: firstly, a yellow sticky trap is installed in the surveillance area to trap flying insects and a camera is set up to collect real-time images. Then the detection and coarse counting method based on You Only Look Once (YOLO) object detection, the classification method and fine counting based on Support Vector Machines (SVM) using global features are designed. Finally, the insect counting and recognition system is implemented on Raspberry PI. Six species of flying insects including bee, fly, mosquito, moth, chafer and fruit fly are selected to assess the effectiveness of the system. Compared with the conventional methods, the test results show promising performance. The average counting accuracy is 92.50% and average classifying accuracy is 90.18% on Raspberry PI. The proposed system is easy-to-use and provides efficient and accurate recognition data, therefore, it can be used for intelligent agriculture applications.
Automatic Buyer Machine for Beverage Waste Linando, Barry; Jembar Jomantara, Muhammad; Atmadja, Wiedjaja
IOP Conference Series: Earth and Environmental Science,
02/2022, Letnik:
998, Številka:
1
Journal Article
Recenzirano
Odprti dostop
Abstract
Automatic Buyer Machine for Beverage Waste is an automatic machine for buying beverage packaging waste using a Raspberry Pi as the main controller and the Convolutional Neural Network (CNN) ...method. By using object detection, the machine can recognize the classification of trained objects. If the object detection is successful in classifying the object, then the object will be accepted by the machine and then directed to the container according to each classification. This experiment uses the Pre-Trained SSDLITE_MOBILENET_V2_COCO model and works well on a Raspberry Pi. The results of the experiment yield an accuracy rate of 95% when the detection model is trained using an augmentation configuration. It was concluded that the system is able to detect beverage packaging objects according to the classification.
Abstract As data centers have increased in size, there has been a need to create clusters out of cheaper, more affordable commodity parts that can easily be replaced upon failure, and that create ...more affordable data centers overall. However, such large clusters are still outside of feasibility for individuals and small businesses. It is a worthwhile exercise to see if much smaller clusters could be created for such applications, and to compare their performance / price measure to that of the previous traditional data centers. For this paper, such a cluster is created using Raspberry Pis which are small-sized, single-board computers. A data sharing model is built in Python using message passing interface (MPI) that ran on the cluster of the four Raspberry Pis. So as to evaluate the performance of the system, some greedy algorithms are created. During the implementation process, previously unknown skills, including how to create a cluster, programming the infrastructure are learnt.
COMPUTER VISION BASED ON RASPBERRY PI SYSTEM Mohanad ABDULHAMID; Otieno ODONDI; Muaayed AL-RAWI
Applied Computer Science (Lublin),
12/2020, Letnik:
16, Številka:
4
Journal Article
Recenzirano
Odprti dostop
The paper focused on designing and developing a Raspberry Pi based system employing a camera which is able to detect and count objects within a target area. Python was the programming language of ...choice for this work. This is because it is a very powerful language, and it is compatible with the Pi. Besides, it lends itself to rapid application development and there are online communities that program Raspberry Pi computer using Python. The results show that the implemented system was able to detect different kinds of objects in a given image. The number of objects were also generated displayed by the system. Also the results show an average efficiency of 90.206 % was determined. The system is therefore seen to be highly reliable.
Rapid screening of pathogenic bacteria contaminated foods is crucial to prevent food poisoning. However, available methods for bacterial detection are still not ready for in-field screening because ...culture is time-consuming; PCR requires complex DNA extraction and ELISA lacks sensitivity. In this study, a microfluidic biosensor was developed for rapid, sensitive and automatic detection of Salmonella using metal-organic framework (MOF) NH2-MIL-101(Fe) with mimic peroxidase activity to amplify biological signal and Raspberry Pi with self-developed App to analyze color image. First, the target bacteria were separated and concentrated with the immune magnetic nanobeads (MNBs), and labeled with the immune MOFs to form MNB-Salmonella-MOF complexes. Then, the complexes were used to catalyze colorless o-phenylenediamine and H2O2 to produce yellow 2,3-diaminophenazine (DAP). Finally, the image of the catalysate was collected under the narrow-band blue light and analyzed using the Raspberry Pi App to determine the bacterial concentration. The experimental results showed that this biosensor was able to detect Salmonella Typhimurium from 1.5 × 101 to 1.5 × 107 CFU/mL in 1 h with the lower detection limit of 14 CFU/mL. The mean recovery for Salmonella in spiked chicken meats was ~112%. This biosensor integrating mixing, separation, labelling and detection onto a single microfluidic chip has demonstrated the merits of automatic operation, fast reaction, less reagent and small size, and is promising for in-field detection of foodborne bacteria.
•An active vibrating mixer was demonstrated with efficient and continuous-flow mixing.•The improved image processing under blue light was verified with higher sensitivity.•Mixing, separation, catalysis and detection were integrated for automatic operation.•This low-cost biosensor could detect Salmonella as low as 14 CFU/mL within 1 h.
The elderly population continues to grow, this population has many chronic conditions, which is why they take multiple pills at the same time, this situation makes it difficult for them to take their ...medications at the appropriate time, which can cause health complications and even death. death. The objective of this research is to improve medication adherence using a medication dispenser that allows authenticating medication consumption accurately with a high level of usability. To do this, we implemented a dispenser with an alarm system that can be configured from a graphical interface, with internet of things (IoT) to remotely monitor the intake of pills and authenticate their consumption through artificial vision that will use the instant messaging system to inform the caregiver about the situation and finally measure the dispenser usability.
Due to the problem of drinking water scarcity in different cities around the world, there are innovative proposals to automate garden irrigation in homes, to reduce drinking water consumption. For ...this research, a sample of 68 inhabitants of the Region of Arequipa - Peru has been surveyed to know the common habits in the irrigation of the gardens. From this data, two systems have been implemented in two average gardens using the Arduino UNO board (integrating with the Ethernet Shield) and the NodeMCU, each proposal integrates soil moisture sensors, water flow sensor, and actuators, such as the solenoid valve and the relay, besides centralizing the information through an IoT System (Home Assistant or Adafruit IO). This has managed to establish a comparison of both, generating a discussion according to the advantages and disadvantages addressed by each proposal and obtaining a saving of potable water in the irrigation of plants.
The voice of each speaker has a unique specific character, influenced by gender, age, emotion, dialect, etc. The use of voice-based gender identification is growing rapidly, such as in the fields of ...security systems, speech recognition, artificial intelligence, etc. However, in speech processing, there are difficulties where the characteristics of the speech signal based on increasing age are difficult to determine accuracy, and there are overlapping fundamental frequency values between males and females. In this research, modeling of a gender identification system based on voice in real-time has been carried out on a Raspberry Pi device. This system is implemented by 2 methods, namely the YIN algorithm and feature extraction of Mel-Frequency Cepstral Coefficient (MFCC). The test results showed that the success of identification in the tuning parameters of scheme two is better than the first scheme by narrowing the overlapping frequency parameters. In the female test data in the closed test, the accuracy is from 98% to 100%, then in the open test starts from 92% to 96%. Meanwhile, the test data for the male closed test increased from 92% to 98%, and the open test started at 90% and rose to 94%. It indicates that the data used in this research is more suitable to use the second scheme parameter tuning to increase the accuracy of the results.
Due to the rise of the Internet of Things, more devices can connect with the Internet. A large amount of data is collected from devices that could be used for different applications. The development ...of hardware equipment for the Internet of Things not only use for industrial but also for smart homes. The smart home covers broad topics, including remote control of home applications, sensing of humans, temperature-controlling air conditions, and security monitors. When we carry out these topics, the human-machine interface is essential for system applications. A gesture recognition system is applied to many real applications. The reason is that the accuracy rate and real factors are complicated. They commonly use of gesture control service in the market is the sensor board of gesture control. The principle is using the electric field to change and determine the gestures. The limitation requires the close operation, and there is a problem of critical point sensitivity. In this paper, we use the gesture control board to combine with gesture image recognition methods to perform the double authentication gesture recognition. Raspberry Pi is the control center to integrate the intelligent light bulb. HUE makes a gesture recognition system. The results explain that the accuracy rate of the gesture recognition proposed is 90%. Meanwhile, it is higher than the SVM method.