In the edge computing and network communication environment, important image data need to be transmitted and stored securely. Under the condition of limited computing resources, it is particularly necessary to design effective and fast image encryption algorithms. One-dimensional (1D) chaotic maps provide an effective solution for real-time image encryption, but most 1D chaotic maps have only one parameter and a narrow chaotic interval, which has the disadvantage of security. In this paper, a new compound 1D chaotic map composed of a logistic map and tent map is proposed. The new system has two system parameters and an arbitrarily large chaotic parameter interval, and its chaotic signal is evenly distributed in the whole value space so it can improve the security in the application of information encryption. Furthermore, based on the new chaotic system, a fast image encryption algorithm is proposed. The algorithm takes the image row (column) as the cyclic encryption unit, and the time overhead is greatly reduced compared with the algorithm taking the pixel as the encryption unit. In addition, the mechanism of intermediate key associated with image content is introduced to improve the ability of the algorithm to resist chosen-plaintext attack and differential attack. Experiments show that the proposed image encryption algorithm has obvious speed advantages and good cryptographic performance, showing its excellent application potential in secure network communication.