To effectively mitigate the potential loss of agricultural production, it is imperative to promptly identify stressors in paddy crops during the booting development stage and implement appropriate measures to address them. This proactive approach aims to minimize both qualitative and quantitative impacts on agricultural output. The visual nature of symptoms has historically required the involvement of human specialists in the detection and classification of stress in traditional paddy crops. This study presents a novel deep convolution neural network (DCNN) framework designed to automate the identification and categorization of various biotic and abiotic stress factors affecting rice crops based on field photographs. The research employs the VGG-16 Convolutional Neural Network (CNN) model, which has undergone prior training, to autonomously categorize images of rice crops experiencing stress during the trunk growing phase. The disciplined models exhibit an common truth of 90.79% when evaluated on the concealed dataset. This outcome provides evidence for the practical viability of employing bottomless knowledge techniques on a dataset comprising 20,000 field photos. These photos depict five distinct varieties of rice crops and are annotated with ten different stress categories, including a category representing healthy or normal conditions. The findings presented in this study have the potential to enhance agricultural and resource management automation systems as well as mobile applications. The aforementioned approach is characterized by subjectivity and a susceptibility to errors, thereby potentially leading to the formulation of unsuitable assessments in the context of stress management.