Orbital angular momentum (OAM), as a new communication multiplexing dimension, has a wide range of applications in the fields of atmospheric laser communication and satellite communication. Here, dual-channel multiband OAM modulation is explored as a means of information transmission. Two-channel multiband signals are used to generate vortex beams with different topological charge values for coherent superposition to produce different light intensity patterns. These patterns are recorded by a charge-coupled device (CCD) camera at the receiving end, where differences in light intensity information denote differences in symbol information. When dual-channel quaternary information is transmitted, 16 different light intensity patterns can be generated. To ensure the differences among light intensity patterns, a coding scheme based on light intensity correlation is proposed. The effect of atmospheric turbulence on light intensity pattern recognition is then explored. The use of a convolutional neural network (CNN) for light intensity information recognition is found to significantly improve the recognition rate of light intensity information under the influence of atmospheric turbulence.