A Self-Predictable Crop Yield Platform (SCYP) Based On Crop Diseases Using Deep Learning

Abstract

This paper proposes a self-predictable crop yield platform (SCYP) based on crop diseases using deep learning that collects weather information (temperature, humidity, sunshine, precipitation, etc.) and farm status information (harvest date, disease information, crop status, ground temperature, etc.), diagnoses crop diseases by using convolutional neural network (CNN), and predicts crop yield based on factors such as climate change, crop diseases, and others by using artificial neural network (ANN). The SCYP consists of an image preprocessing module (IPM) to determine crop diseases through the Google Vision API and image resizing, a crop disease diagnosis module (CDDM) based on CNN to diagnose the types and extent of crop diseases through photographs, and a crop yield prediction module (CYPM) based on ANN by using information of crop diseases, remaining time until harvest (based on the date), current temperature, humidity and precipitation (amount of snowfall) in the area, sunshine amount, ground temperature, atmospheric pressure, moisture evaporation in the ground, etc. Four experiments were conducted to verify the efficiency of the SCYP. In the CDMM, the accuracy and operation time of each model were measured using three neural network models: CNN, region-CNN(R-CNN), and you only look once (YOLO). In the CYPM, rectified linear unit (ReLU), Sigmoid, and Step activation functions were compared to measure ANN accuracy. The accuracy of CNN was about 3.5% higher than that of R-CNN and about 5.4% higher than that of YOLO. The operation time of CNN was about 37 s less than that of R-CNN and about 72 s less than that of YOLO. The CDDM had slightly less operation time, but in this paper, we prefer accuracy over operation time to diagnose crop diseases efficiently and accurately. When the activation function of the ANN used in the CYPM was ReLU, the accuracy of the ANN was 2% higher than that of Sigmoid and 7% higher than that of Step. The CYPM prediction was about 34% more accurate when using multiple diseases than when not using them. Therefore, the SCYP can predict farm yields more accurately than traditional methods.

A Systematic Review of IoT Solutions for Smart Farming

“…Similarly, [ 73 ] uses fuzzy logic to optimize the number of sensors for monitoring soil temperature and moisture. Machine learning was also used in data processing by [ 46 ] to predict environmental conditions based on the forecast values of weather, humidity, temperature and water level and thus to control an irrigation system, by [ 47 ] to combine multiple parameters obtained from images, such as color and texture indices and by [ 48 ] to identify marks on the plants and, thus, to identify possible diseases. Similarly, in [ 58 , 125 ] it was used to detect diseases, identify growth stages and the health of plantations.…”

“…Similarly, as shown in Table 7 , IoT solutions used computer vision for applications that need to deal with image processing, such as crop monitoring and diseases prevention. It was also possible to observe in the reviewed papers the use of computer vision to identify and classify elements in images obtained by cameras, enabling the identification of fruit in an orchard [ 200 ] or the existence of diseases and pests in plantations [ 48 , 129 , 133 ]. Additionally, in [ 133 ] computer vision was used as a monitoring tool to detect the presence of insects that can cause diseases in olive groves and in [ 48 ] the same technique was employed to analyze diseases that cause morphological deformations in plants.…”

“…It was also possible to observe in the reviewed papers the use of computer vision to identify and classify elements in images obtained by cameras, enabling the identification of fruit in an orchard [ 200 ] or the existence of diseases and pests in plantations [ 48 , 129 , 133 ]. Additionally, in [ 133 ] computer vision was used as a monitoring tool to detect the presence of insects that can cause diseases in olive groves and in [ 48 ] the same technique was employed to analyze diseases that cause morphological deformations in plants. Additionally, computer vision was used in crop management systems, for example in [ 134 ] where it was implanted in a robot equipped with a camera and other sensors, being able to obtain images of vegetation and, through computer vision, detect weeds in plantations and eliminate them.…”

“…In this context, big data allows distributed storage and parallel data processing, enabling the extraction of information in the shortest possible time [ 43 ]. Such information are used as models by artificial intelligence (AI) systems—which, according to [ 44 ], can be understood as the ability of a system to operate as if it had the thinking capacity of a human being—and machine learning—that, according to [ 45 ] is a data processing technique to detect patterns and correlation among complex and unrelated data—for the development of decision support systems and automation of irrigation control systems [ 46 ], monitoring [ 47 ] and diseases detection in crops [ 48 ], for example.…”

MDFC–ResNet: An Agricultural IoT System to Accurately Recognize Crop Diseases

“…Savary [3] estimated global losses for wheat at 21.5%, rice at 30%, maize at 22.5%, potatoes at 17.2%, and soybeans at 21.4%. Generally, crop losses due to pathogens, animals, and weeds are approximately between 20% and 40% of global production [2], [4]- [6]. In the near future, global warming may increase crop losses through more active fungi [7] and insects [8].…”