Industrial quality inspection aims to identify defective parts in industrial production processes. Commonly used methods for industrial quality inspection rely on feature representations that have been pre-trained on natural image datasets, such as ImageNet. However, these pre-trained models are not specifically tailored for industrial scenarios and therefore do not transfer well to downstream industrial tasks. In this study, we have curated a large-scale industrial production dataset called Ind-2M, which is specifically collected from industrial scenarios. This dataset serves to enhance the industrial representation of pre-training models. Additionally, we propose a Pixel-level COntrastive pre-trainer (PiCO) for industrial image representation. PiCO not only improves the global industrial representation through industrial production classification, but also enhances the local industrial representation through pixel-level self-supervision. Experimental results demonstrate that PiCO effectively transfers to downstream industrial tasks, such as multi-label defect classification and anomaly detection, outperforming existing pre-trained methods. We hope PiCO can initiate a new paradigm for industrial image pre-training.