YOLO‐DFD: A Lightweight Method for Dog Feces Detection Based on Improved YOLOv4

Abstract

Computer vision has been integrated into people’s daily lives, but mainstream target detection algorithms deployed to embedded devices with limited hardware resources are difficult to meet the task requirements in terms of real time and accuracy. So we proposed YOLO-DFD, a light detection algorithm based on improved YOLOv4 to solve the problem of dog feces in our living environment. The main improvement strategies are as follows: the YOLOv4 backbone network is replaced with MobileNetV3, and the 3 ∗ 3 convolutions in the feature enhancement network are replaced by depthwise separable convolutions to further reduce the number of parameters. To enhance the accuracy of target detection, we introduced the convolutional block attention module (CBAM) in neck network, and the complete intersection over union (CIoU) loss of YOLOv4 is replaced with the SCYLLA intersection over union (SIoU) loss to reduce false detection rate. In this paper, the dataset we used was made up of pictures of dog feces taken in life. For the self-made dog feces dataset, we used data enhancement technology to expand it. The training result shows that the average precision (AP) has reached 98.66%. While maintaining detection performance, the parameter of YOLO-DFD is reduced by 82% and FPS increases 14 compared to YOLOv4. And YOLO-DFD has a lower parameter quantity and a smaller calculation than other algorithms, making it easier to deploy on embedded devices to clean dog feces.

Research on pavement crack detection algorithm based on improved Yolov5

“…5, and then these two feature maps are channel spliced and transformed into 1-channel weight maps after 7*7 convolution operation, and generates the spatially compressed weight matrix M s by sigmoid. Finally, M s and the input 𝐅 ′ of the module are multiplied to get the final generated features 𝐻 ′ * 𝑊 ′ * 1 [15]. The computational formula is:…”

Catfish Fry Detection and Counting Using YOLO Algorithm

“…Penelitian yang berhubungan, yaitu menghitung orang dan sepeda [4] menyimpulkan bahwa sistem YOLO mampu mendeteksi dan menghitung objek pengendara sepeda secara real-time dengan 18 kesalahan perhitungan pada 525 objek dan waktu inferensi rata-rata 112,82 ms per frame. Penelitian mendeteksi kotoran anjing dengan memodifikasi algoritma sehingga mendapatkan akurasi yang lebih baik mencapai 98,66% [6]. Penelitian [7] yang sama juga dilakukan menggunakan teknologi YOLO untuk mendeteksi dan mengklasifikasi jenis kapal yang melintas di dermaga dengan kinerja 97,59% dan dapat secara akurat menemukan kapal dilingkungan kegelapan mencapai 96,13% sehingga mencapai deteksi dan klasifikasi kapal lebih efektif.…”

Machine learning based blueberry detection method by CIE-YOLOv5

“…Even in complex environment, the final average means of accuracy by the YOLOv3 model is 92.27% [5]. The problem with computational complexity of YOLO is improved by a lightweight detection method, which can detect apple target in real-time with 92.23% accuracy [6]. However, the original YOLOv5 model has poor stability and robustness for the recognition and detection of blueberries, especially complex natural environment.…”