Translating C code into Rust is an effective way to ensure its memory safety. Rule-based translation produces Rust code that remains unsafe and fails to leverage Rust’s safety features. LLM-based methods can generate more idiomatic and safer Rust as LLMs are trained on large amounts of human-written code. Although promising, existing LLM-based methods still struggle with project-level C-to-Rust translation. They typically partition a C project into smaller units (e.g., functions) based on the call graph and translate them bottom-up to resolve dependencies. However, this bottom-up, unit-by-unit paradigm often fails to handle pointer translation due to the lack of a global perspective on pointer usage. To address this, we propose a novel C-Rust Pointer Knowledge Graph (KG) that enriches a code-dependency graph with two types of pointer semantics: (i) pointer-usage information, recording global behaviors such as points-to flows and lower-level struct usage by higher-level units; and (ii) Rust-oriented annotations, encoding ownership, mutability, nullability, and lifetimes. Based on the C-Rust Pointer KG and its synergy with LLMs, we further propose PtrTrans, a project-level C-to-Rust translation technique. In PtrTrans, the C-Rust Pointer KG provides LLMs with comprehensive pointer semantics from a global perspective, thus guiding LLMs towards generating safe and idiomatic Rust code from a given C project. Our experiments show that PtrTrans reduces unsafe usages in translated Rust by 99.9% compared to both rule-based translation and LLM-based rewriting, while achieving an average 29.3% higher functional correctness than fuzzing-enhanced LLM methods.