In this paper, a fast smoke image detection algorithm is proposed to extract smoke features of fire. The irregular shapes and color variation of smoke, makes it challenging to detect smoke accurately. The situation is even more complicated in complex sites like forests, where traditional smoke sensors are hard to install and have difficulty detecting smoke. Hence, the power of image processing is utilized. We propose a parallelized smoke image detection algorithm for the original algorithm in [1]. The new algorithm relies on dividing the received images into segments and analyzing these segments using multi-threading techniques. Parallelism is utilized at both the macro and micro levels. One thread is created to handle each received image from the area under surveillance. Concurrently, multiple threads are generated to analyze each image independently from the other received images. Various detection experiments are conducted, and the results show that the proposed algorithm is able to enhance the detection speed without compromising the detection accuracy. Compared with the original algorithm, the proposed algorithm operates with higher speed, making it suitable for monitoring more complex and secure environments.