Currently, data processing computing systems primarily rely on the Von Neumann architecture. This architecture employs serial data processing and physically separates the processor unit from the storage unit, resulting in data transmission delays. These delays not only reduce work efficiency but also increase energy consumption. Neuromorphic computing has gained significant attention for its ability to process large amounts of data with minimal power consumption. Artificial neurons are crucial components of this technology and have been extensively researched. The primary function of these devices is to receive and integrate input signals from synapses and generate spike signals as outputs when the threshold is exceeded. Artificial neurons typically use a threshold function to determine whether synaptic signals are integrated enough to reach the threshold. They receive postsynaptic current from the previous synapse as input and output voltage in a spike form to the front end of the next synapse, firing like biological neurons to exchange information. Researching more efficient and precise artificial neural devices is of great significance for processing complex information. Therefore, it is important to continue exploring the potential of memristor-based artificial neurons. Artificial neurons based on memristors have advantages such as high stacking density, low power consumption, and fast switching speed, which are essentially closer to biological neurons. Currently, artificial neurons based on memristors are mainly categorised into three types: electrochemical mechanism-based, valence mechanism-based, and phase change mechanism-based. To process complex information more efficiently in artificial neural morphology computing, we propose an artificial neuron device based on Ag/IDTBT/ZnO/Si memristors. The device exhibits good threshold characteristics, with a switching ratio of about 10²~10³ and lower operating voltage. It can simulate a neuron model for Leaky Integrate and Fired ignition, with the ignition time of the neuron device being inversely proportional to the pulse amplitude applied to the device. Increasing the applied pulse amplitude from 0.8 V to 1 V results in a decrease in the integrated ignition time of the device from 5.22 s to 1.19 s. The ignition time decreases as the amplitude increases. It is important to note that if the applied pulse amplitude is too low, the neuron device cannot be activated, while if it is too high, the device irreversibly breaks down and the internal lattice structure of the material is permanently damaged. In complex neural morphology calculations, artificial neurons require adjustable performance to adapt to their environment. Therefore, we investigated the impact of Indacenodithiophene-co-benzothiadiazole (IDTBT) concentration on the performance of artificial neural devices. The results indicate that an increase in IDTBT concentration can lead to an increase in film thickness. This, in turn, can increase the threshold voltage of the neural device and the amplitude voltage required for integral ignition. Currently, most artificial neurons are driven by electrical signals. However, these signals have some drawbacks, such as high power consumption, limited triggering selection, and difficulty in simulating visual systems, which hinder further improvements in computing speed and energy efficiency. In contrast, optical signals offer significant advantages in terms of high energy efficiency, high bandwidth, low crosstalk, and computational speed. To enhance the operating speed of the neural morphology system, we investigated the photoelectric synergistic effect of photoelectric neural morphology devices and the impact of light on device performance. Upon illumination, the device's threshold voltage decreased significantly from 1.99 V to 1.62 V. To assess the device's stability, we retested it after 30 days of storage. By comparing the switch ratio and threshold voltage of two time periods, it appears that the switch ratio and threshold voltage remain stable at approximately 10³ and 1.79 V, respectively. The device's overall performance is stable without significant changes, indicating good stability. This work presents an effective strategy for promoting the development of the neuromorphic system.