A novel automated cloud-based image datasets for high throughput phenotyping in weed classification

Abstract

Deep learning-based weed detection data management involves data acquisition, data labeling, model development, and model evaluation phases. Out of these data management phases, data acquisition and data labeling are labor-intensive and time-consuming steps for building robust models. In addition, low temporal variation of crop and weed in the datasets is one of the limiting factors for effective weed detection model development. This article describes the cloud-based automatic data acquisition system (CADAS) to capture the weed and crop images in fixed time intervals to take plant growth stages into account for weed identification. The CADAS was developed by integrating fifteen digital cameras in the visible spectrum with gphoto2 libraries, external storage, cloud storage, and a computer with Linux operating system. Dataset from CADAS system contain six weed species and eight crop species for weed and crop detection. A dataset of 2000 images per weed and crop species was publicly released. Raw RGB images underwent a cropping process guided by bounding box annotations to generate individual JPG images for crop and weed instances. In addition to cropped image 200 raw images with label files were released publicly. This dataset hold potential for investigating challenges in deep learning-based weed and crop detection in agricultural settings. Additionally, this data could be used by researcher along with field data to boost the model performance by reducing data imbalance problem.

Keywords: Automated data acquisition; Cloud computing; Computer vision; Deep learning; Weed and crop detection.

Fig. 1

Folder structure of the data repository that is publicly available.

Fig. 2

Automated image acquisition system components: a) Canon EOS Rebel 7T, b) Canon EOS 90D, c) Desktop computer with Linux OS, d) AC battery adapters for continuous power, e) LCD display f) Data storage disk g) USB hubs and h) power extension cords.

Fig. 3

Block diagram showing automated image acquisition system with its component; external storage device, cloud, desktop with ubuntu 20.04, and canon camera; arrow directions show the direction of image flow from the camera.

Fig. 4

Automated image acquisition system greenhouse setup: a) Greenhouse benches with weed and crop pots, b) digital camera, c) Desktop computer with Linux OS attached to an external storage drive.

Fig. 5

Flowchart showing the Linux bash script algorithm for automatic image acquisition.

Fig. 6

Flowchart illustrating the process of extracting a cropped image containing six weeds and eight crops from the original image.

Fig. 7

Flowchart depicting the methodology for cropping raw images to extract individual object classes, including relevant mathematical formulations.